CN101140619A - Image processing device, data processing device and parameter adjusting method - Google Patents
Image processing device, data processing device and parameter adjusting method Download PDFInfo
- Publication number
- CN101140619A CN101140619A CNA2007101458854A CN200710145885A CN101140619A CN 101140619 A CN101140619 A CN 101140619A CN A2007101458854 A CNA2007101458854 A CN A2007101458854A CN 200710145885 A CN200710145885 A CN 200710145885A CN 101140619 A CN101140619 A CN 101140619A
- Authority
- CN
- China
- Prior art keywords
- parameter
- value
- aforementioned
- data
- image
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Images
Landscapes
- Image Processing (AREA)
- Image Analysis (AREA)
Abstract
The present invention provides a data processing device, an image processing apparatus and a parameter regulation means, capable of reliably regulating parameters related by a data processing through easy manipulation. When a regulation of parameter values related by an image processing is executed, a target data creation part (140) receives revising dictation of result data D2 from a manipulator to create a target data D3. Then, a data regulating and processing part (170) detects the parameter values of the result data D2, which can be obtained, are consistent with the parameter values of the target data D3 within a specified permissible range and are then determined as the adjustment values.
Description
Technical field
The present invention relates to adjust the technology of the parameter relevant (for example, handling relevant parameter) with defects detection with various data processing (for example, Flame Image Process).
Background technology
In the printing work flow, check whether produce unexpected defective etc. (hereinafter referred to as " examining the inspection of version/printed matter ") in the graphic arts film made, brush version, the printed matter etc.This inspection version/printed matter inspection is performed such, that is, will be as the image of checking object (specifically, scanning obtain data) as the printed matter of checking object etc., compare with the view data that generates the benchmark in source as it, detect its difference as defective.
In inspection version/printed matter is checked, by various processing (for example, Fuzzy Processing, rock that processing, contrast are handled, isolated point removes and handles etc.), the defective that detection produces on as the image of checking object, but the detection sensitivity of defective can change according to the value of the environmental parameter of each processing.That is to say,, need suitably adjust parameter value in order to carry out defects detection with appropriate sensitivity.
The operation of this parameter adjustment, current is to be undertaken by the method for trial and error that each user revises with empirical method.
In addition, handle relevant prior art, propose following technology as checking with the inspection version, for example, the printed data that will become inspection version object according to characteristics of image carries out subregion, in each zone, the attribute corresponding parameter of use and this image-region is examined version and is handled (referring to Patent Document 1).Under this situation,, also need suitably to adjust in advance parameter value in order in each image-region, to carry out suitable defects detection.
Patent documentation 1: the spy opens the 2004-258470 communique
Summary of the invention
In inspection version/printed matter is checked, carry out above-mentioned various processing with mixing, the data class that should adjust is also a lot.Therefore, in parameter adjustment, must expend plenty of time and work.
In addition, inferior even with having adjusted under the situation of parameter someway in the situation of the detection sensitivity that needs the flip-flop defective, because do not readjust the plenty of time of parameter value, so finally can only check by range estimation.
The present invention proposes in view of above-mentioned problem, its purpose is to provide a kind of data processing equipment, image processing apparatus and parameter regulation means, it can enough shirtsleeve operations, adjust exactly that data processing (for example, Flame Image Process) relates to more than or equal to 1 parameter.
The invention of technical scheme 1 has: graphics processing unit, the Flame Image Process that it is stipulated; The target image generation unit, it receives the revision directive to the result images that generates as aforementioned processing result image, generate make aforementioned revision directive anti-, in the aforementioned result image and the target image that obtains; And the parameter adjustment unit, it adjusts the value of the parameter relevant with aforementioned Flame Image Process, so that the result images that generates as aforementioned processing result image is consistent with aforementioned target image.
The invention of technical scheme 2 is, as technical scheme 1 described image processing apparatus, aforementioned graphics processing unit carries out following Flame Image Process, promptly, the inspection object images that benchmark image is obtained with this benchmark image of output compares, and detects the defect part that produces in aforementioned inspection object images, generates the result images that shows this detected defect part with the display format different with non-defect part, the parameter relevant with aforementioned Flame Image Process, the detection sensitivity of regulation aforementioned disadvantages part.
The invention of technical scheme 3 is, as technical scheme 1 or technical scheme 2 described image processing apparatus, aforementioned target image generation unit is presented at the aforementioned result image in the display frame, from the input of aforementioned display frame reception to the revision directive of this result images.
The invention of technical scheme 4 is, as technical scheme 3 described image processing apparatus, the aforementioned parameters adjustment unit makes the value of aforementioned parameters change on one side, judge that on one side whether the result images that obtains is consistent with aforementioned target image, is defined as the adjusted value of aforementioned parameters can obtain the parameter value of the result images consistent with aforementioned target image under each parameter value.
The invention of technical scheme 5 is, as technical scheme 4 described image processing apparatus, the aforementioned parameters adjustment unit is defined as the median of following set with the adjusted value of aforementioned parameters, and this set is made of the parameter value that can obtain the result images consistent with aforementioned target image.
The invention of technical scheme 6 is, as technical scheme 4 described image processing apparatus, the aforementioned parameters adjustment unit, the aforementioned parameters adjusted value is defined as following parameter value, this parameter value is the parameter value in the time of can obtaining the result images consistent with aforementioned target image at first in the process that the aforementioned parameters value is changed.
The invention of technical scheme 7 is, as technical scheme 4 described image processing apparatus, also has the setting range designating unit, and it specifies the scope that the aforementioned parameters value is changed according to the setting value of input.
The invention of technical scheme 8 is, as technical scheme 4 described image processing apparatus, also has the variable quantity designating unit, and it specifies the variable quantity when the aforementioned parameters value is changed according to the setting value of input.
The invention of technical scheme 9 is, as technical scheme 3 described image processing apparatus, the aforementioned parameters adjustment unit is from the combination of the parameter value more than 1 or 1, utilize genetic algorithm, extraction can obtain the parameter value combination of the result images consistent with aforementioned target image, the adjusted value of aforementioned parameters is defined as constituting each value of this combination that extracts.
The invention of technical scheme 10 is, as technical scheme 9 described image processing apparatus, also has the setting range designating unit, and it specifies the scope that the aforementioned parameters value is changed according to the setting value of input.
The invention of technical scheme 11 is, as technical scheme 9 described image processing apparatus, also has the variable quantity designating unit, and it specifies the variable quantity when the aforementioned parameters value is changed according to the setting value of input.
The invention of technical scheme 12 is, as technical scheme 1 or 2 described image processing apparatus, have the image parameter of adjustment designating unit, it is according to the characteristic quantity of the regulation that obtains from aforementioned revision directive, from the parameter that aforementioned Flame Image Process relates to, specifying becomes the parameter of adjusting object.
The invention of technical scheme 13 has: target image generates operation, it receives the revision directive of the result images that generates to becoming the result who carries out Flame Image Process from the operator, generates aforementioned revision directive to be reflected in the aforementioned result image and the target image that obtains; And the parameter adjustment operation, it adjusts the value of the parameter relevant with aforementioned Flame Image Process, so that the result images that generates as carrying out the result of aforementioned Flame Image Process is consistent with aforementioned target image.
The invention of technical scheme 14 has: data processing unit, and it carries out predetermined data and handles; The target data generation unit, it receives the revision directive of the result data that the result who handles as aforementioned data is obtained, and generation makes aforementioned revision directive be reflected in the target data that obtains in the aforementioned result data; And the parameter adjustment unit, it adjusts the value of handling relevant parameter with aforementioned data, so that the data of obtaining as the aforementioned data result are consistent with aforementioned target data.
The invention of technical scheme 15 is, as technical scheme 14 described data processing equipments, the aforementioned data processing unit carries out following Flame Image Process, promptly, the inspection object images that benchmark image is obtained with this benchmark image of output compares, the defect part that detection produces in aforementioned inspection object images, generation shows the view data of this detected defect part with the display format different with non-defect part, obtain as the aforementioned result data, handle relevant parameter with aforementioned data, the detection sensitivity of regulation aforementioned disadvantages part.
The invention of technical scheme 16 is, as technical scheme 14 or technical scheme 15 described data processing equipments, aforementioned target data generation unit in display frame, receives input to the revision directive of this result data from aforementioned display frame with the aforementioned result data presentation.
The invention of technical scheme 17 is, as technical scheme 16 described data processing equipments, the aforementioned parameters adjustment unit makes the value of aforementioned parameters change on one side, judge on one side whether the result data that obtains is consistent with aforementioned target data under each parameter value, with the adjusted value of aforementioned parameters, be defined as to obtain the parameter value of the result data consistent with aforementioned target data.
The invention of technical scheme 18 is, as technical scheme 17 described data processing equipments, the aforementioned parameters adjustment unit is defined as the median of following set with the adjusted value of aforementioned parameters, and this set is made of the parameter value that can obtain the result data consistent with aforementioned target data.
The invention of technical scheme 19 is, as technical scheme 17 described data processing equipments, the aforementioned parameters adjustment unit, the adjusted value of aforementioned parameters is defined as following parameter value, and this parameter value is the parameter value when obtaining the result data consistent with aforementioned target data at first in the process that the aforementioned parameters value is changed.
The invention of technical scheme 20 is, as technical scheme 17 described data processing equipments, also has the setting range designating unit, and it specifies the scope that the aforementioned parameters value is changed according to the setting value of input.
The invention of technical scheme 21 is, as technical scheme 17 described data processing equipments, also has the variable quantity designating unit, and it specifies the variable quantity when the aforementioned parameters value is changed according to the setting value of input.
The invention of technical scheme 22 is, as technical scheme 16 described data processing equipments, the aforementioned parameters adjustment unit, from the combination of the parameter value more than 1 or 1, utilize genetic algorithm, extraction can draw the parameter value combination of the result data consistent with aforementioned target data, the adjusted value of aforementioned parameters is defined as constituting each value of this combination that extracts.
The invention of technical scheme 23 is, as technical scheme 22 described data processing equipments, also has the setting range designating unit, and it specifies the scope that the aforementioned parameters value is changed according to the setting value of input.
The invention of technical scheme 24 is, as technical scheme 22 described data processing equipments, also has the variable quantity designating unit, and it specifies the variable quantity when the aforementioned parameters value is changed according to the setting value of input.
The invention of technical scheme 25 is, as technical scheme 14 or 15 described data processing equipments, have the image parameter of adjustment designating unit, it is according to the characteristic quantity of the regulation of obtaining from aforementioned revision directive, from handling the relevant parameter, specify as the parameter of adjusting object with aforementioned data.
The invention of technical scheme 26 has:
Target data generates operation, and it receives revision directive to the result data that obtains as carrying out The results of data processing from the operator, generate make aforementioned revision directive anti-, in the aforementioned result data and the target data that obtains; And the parameter adjustment operation, it adjusts the value of handling relevant parameter with aforementioned data, so that the data of obtaining as the result who carries out the aforementioned data processing are consistent with aforementioned target data.
The effect of invention
According to technical scheme 1,13 described inventions, the operator is not direct input parameter value, but by providing the revision directive to result images, carries out parameter adjustment.That is to say, can use shirtsleeve operation, adjust the parameter relevant with Flame Image Process.In addition, make the image parameter value consistent that obtains as processing result image because be adjusted into, so can obtain the parameter value of operator's desired images result with target image.That is to say, can adjust the parameter that Flame Image Process relates to exactly.
According to technical scheme 2,15 described inventions, can be easily and adjust the parameter that regulation is checked the detection sensitivity of the defect part that produces in the object images exactly.
According to technical scheme 3 described inventions, because receive input, so the operator can be easily and imported revision directive to result images exactly to the revision directive of result images from display frame.Consequently, can be more prone to and carry out parameter adjustment exactly.
According to technical scheme 4 described inventions, because make parameter value variation, and meanwhile detect suitable parameter value (that is, making the result images parameter value consistent that under this parameter, obtains) with target image, so can detect suitable parameter value exactly.
According to technical scheme 5 described inventions, because the parameter adjustment value is defined as the median of following set, that is, and the set that constitutes by the parameter value that can obtain the result images consistent with target image, so in suitable parameter value, can determine that also optimum value is as the parameter adjustment value.
According to the invention of technical scheme 6, because the parameter adjustment value is defined as following parameter value, this parameter value is for making that in the process of parameter value variation, the parameter value when obtaining the result images consistent with target image at first is so can carry out parameter adjustment rapidly.
According to technical scheme 9,22 described inventions,, specify the scope of parameter value variation that makes, so can dwindle the setting range of parameter because according to the setting value of importing.Thus, can carry out parameter adjustment rapidly.
According to technical scheme 7,10,20,23 described inventions,, specify and make the scope of parameter value variation, so can dwindle the setting range of parameter because according to the setting value of importing.Thus, can carry out parameter adjustment rapidly.
According to technical scheme 8,11,21,24 described inventions, because according to the setting value of importing, specify the variable quantity when making parameter value variation, so, can reduce the parameter value that can in the setting range of parameter, obtain.Thus, can carry out parameter adjustment rapidly.
According to technical scheme 12 described inventions, because according to the characteristic quantity of the regulation of from revision directive, obtaining, specify and adjust image parameter, so, can suitably select to wish to be adjusted into the kind of the parameter that obtains the result images consistent with target image.Thus, can be rapidly and carry out parameter adjustment exactly.
According to technical scheme 14,26 described inventions, the operator can not be direct input parameter value, but by providing parameter adjustment is carried out in the revision directive of result data.That is to say, can adjust the parameter that data processing relates to shirtsleeve operation.In addition, make the data parameter consistent that obtains as data processed result because be adjusted into, so obtain having the parameter value of operator's desired data result with target data.That is to say, can adjust the parameter that data processing relates to exactly.
According to technical scheme 16 described inventions, because receive input, so the operator can be easily and imported revision directive to result data exactly to the revision directive of result data from display frame.Consequently, can be more prone to and carry out parameter adjustment exactly.According to technical scheme 17 described inventions, because make parameter value variation, and meanwhile detect suitable parameter value (that is, the parameter value that the result data that obtains is consistent with target data) under this parameter value, so can detect suitable parameter value exactly.
According to technical scheme 18 described inventions, because the parameter adjustment value is defined as the median of following set, that is, and the set that constitutes by the parameter value that can obtain the result data consistent with target data, so in suitable parameter value, can determine that also optimum value is as the parameter adjustment value.
According to technical scheme 19 described inventions, because with the parameter adjustment value, be defined as following parameter value, that is, making in the process of parameter value variation, obtain the parameter value of the result data consistent at first with target data.
According to technical scheme 25 described inventions, because according to the regulation characteristic quantity that from revision directive, obtains, specify and adjust image parameter, so, can suitably select to be adjusted into the parameter kind that obtains the result data consistent with target data.Thus, can be rapidly and carry out parameter adjustment exactly.
Description of drawings
Fig. 1 is the structural drawing of print system 100 of the flaw detection apparatus 1 of an expression mode containing the image processing apparatus that embodiments of the present invention 1 relate to.
Fig. 2 is the explanation structural drawing relevant with the parameter adjustment function.
Fig. 3 is illustration reference data D0 and the figure that checks object data D1
Fig. 4 is the figure of illustration result data D2 and target data D3.
Fig. 5 is the figure of the whole flow process of the parameter adjustment processing in the expression flaw detection apparatus 1.
Fig. 6 is the figure of the flow process handled of the generation of expression target data D3.
Fig. 7 is the figure that expression receives the structure example of picture G.
Fig. 8 is the figure of the flow process of the expression designated treatment of adjusting image parameter.
Fig. 9 is the figure that expression receives the structure example of picture H.
Figure 10 is the figure that the expression parameter value is determined treatment scheme.
Figure 11 is used to illustrate that the median of the tentative calculation value that will be judged as appropriate value is chosen as the figure of the situation of adjusted value.
Figure 12 is the explanation structural drawing relevant with the parameter adjustment function.
Figure 13 is the figure of illustration result data D2 and target data D3.
Figure 14 is the figure that the illustration amount of brightness obtains the zone.
Figure 15 is the figure of the histogram shape of illustration Luminance Distribution.
Figure 16 is the figure that the flow process of histogram shape kind is judged in expression.
Figure 17 is the figure that the flow process of image parameter designated treatment is adjusted in expression.
Figure 18 is that illustration is stored in the figure that adjusts the judged result in the image parameter determination portion 253.
Figure 19 is the figure of definite treatment scheme of expression parameter value.
Figure 20 is the explanation structural drawing relevant with the parameter adjustment function.
Figure 21 is the concept map that definite mode of the adjusted value that has used genetic algorithm is described.
Figure 22 is the concept map that definite mode of the adjusted value that has used genetic algorithm is described.
Figure 23 is the figure of definite treatment scheme of expression adjusted value.
Embodiment
(the 1st embodiment)
＜1. structure 〉
Fig. 1 is the structural drawing of print system 100 of the flaw detection apparatus 1 of an expression mode containing the data processing equipment that relates to as the 1st embodiment of the present invention.Flaw detection apparatus 1 specifies as described later, is the device that detects the defective that produces in the printed matter that generates etc., has the function of adjusting the Flame Image Process parameter that this defects detection processing relates to.
＜1-1. print system structure 〉
Print data generation apparatus 2 is the devices that generate printed data.More particularly, it carries out the layout processing such as image configurations in the printed images and generates topology data, and, handle (grating processings) by RIP, the topology data of generation is transformed to many image gray data.In addition, handle, also can adopt the structure of in the autonomous device different, carrying out with print data generation apparatus 2 for RIP.
Print data generation apparatus 2 can generate the view data of the various resolutions corresponding with application target according to same topology data.For example, can generate the site view data of the high-res of 2400dpi degree, as the view data that output is used, on the other hand, the multi-grey image data of the low-res of output 300dpi degree are as the reference data D0 that checks usefulness.But the resolution of reference data D0 is preferably according to checking that object data D1 determines that in the present embodiment, the resolution of reference data D0 is set at and the level of checking that object data D1 is identical.For example, take printed matter at the digital camera with the 300dpi resolution, obtain to check under the situation of data object D1, print data generation apparatus 2 is carried out RIP to topology data and is launched with the resolution of 300dpi, generates reference data D0.The reference data D0 that generates via network N, sends flaw detection apparatus 1 to, is used for the inspection of inspection version/printed matter.
The structure of＜1-2. flaw detection apparatus 〉
In addition, in flaw detection apparatus 1, can realize so-called GUI (Graphical User Interface) 101 (with reference to Fig. 2) by the function of control part 11, operating portion 13, display part 14, promptly, can make the content of operation that undertaken by operating portion 13 or, handle while be presented on the display part 14 for treatment situation of various processing etc.Thus, the operator can be by one side with reference to the picture that is presented on the display part 14, Yi Bian the operation of using operating portion 13 to stipulate is sent instruction to computing machine.
The structure that＜1-3. is relevant with parameter adjustment 〉
Reference data acquisition unit 110 according to the operator's who receives via operating portion 13 instruction, is obtained reference data D0.More particularly, via network N, obtain reference data D0 from print data generation apparatus 2.In addition, as shown in Figure 1, can also utilize the media read/write device 151 that constitutes by MO driver or CD-R/RW driver etc., read the reference data D0 in the recording medium M (for example MO (photomagneto disk) or CD-R/RW etc.) that is stored in mobility, thereby obtain reference data D0.Reference data acquisition unit 110 is stored in the reference data D0 that obtains in the storage part 12.
Check object data acquisition unit 120,, obtain and check object data D1 according to the operator's who receives via operating portion 13 instruction.More particularly, via network N, obtain inspection object data D1 by image analyzer 5.In addition, also can read the recording medium M of having stored the inspection object data D1 that obtains with shootings such as digital camera inspection object, check object data D1 thereby obtain by media read/write device 151.Check that the inspection object data D1 that object data acquisition unit 120 will be obtained is stored in the storage part 12.
Illustration reference data D0 in Fig. 3 (a).In addition, in Fig. 3 (b), the inspection object data D1 of the inspection object that illustration generates according to the illustrative reference data D0 of Fig. 3 (a).As mentioned above, reference data D0 is the view data as the benchmark that generates printed matter, checks that object data D1 is a view data of checking object.Originally, checked that object data D1 should be in full accord with reference data D0.But,, as shown in Figure 3, in checking object data D1, may produce the part (defective) different with reference data D0 because the various factors that in printing process, takes place (for example, dust is sneaked into, position deviation).Defects detection handling part 130 described later by carrying out defects detection algorithm described later, detects this defective.
Defects detection handling part 130 carries out following Flame Image Process, promptly, to being stored in the reference data D0 in the storage part 12, the inspection object data D1 that obtains with output this reference data D0 compares, the defect part that detection produces in checking object data D1 generates the result data D2 that shows this detected defect part with the display mode different with non-defect part.More particularly, for being stored in the reference data D0 in the storage part 12 and checking object data D1, the defects detection that puts rules into practice detects the defective that produces in checking object data D1.So-called defects detection Flame Image Process, more particularly, be meant Fuzzy Processing, rock processing, contrast is handled, isolated point is removed and handled, these Flame Image Process are at the series of algorithms that is used for detecting defective (hereinafter referred to as " defects detection algorithm "), order is according to the rules carried out.
Defects detection handling part 130 has the handling part that carries out the defects detection Flame Image Process, promptly Fuzzy Processing portion 131, rock handling part 132, comparison process portion 133 and isolated point is removed handling part 134.In addition, have result images generating unit 135, it reflects testing result and generates result data D2.
Isolated point is removed handling part 134, carries out the isolated point removal according to the parameter (isolated point is removed parameter P4) relevant with isolated point removal processing and handles.So-called isolated point is removed and is handled, and is meant under reference data D0 and the pixel of checking generation difference between object data D1 are isolated to situation in the less pixel coverage, with its processing of removing as unwanted isolated point.Being judged as isolated pixel coverage (regulation of more specifically saying so forms the isolated point of the number of pixels of isolated point and removes pixel count) here, sets as isolated point and removes parameter P4.
In addition, the defects detection algorithm in the present embodiment is by this each handling part 131～134, and to the reference data D0, the inspection object data D1 that read from storage part 12, the processing that puts rules into practice is carried out, and more particularly, carries out according to following flow process.
At first, Fuzzy Processing portion 131 carries out Fuzzy Processing to reference data D0 and inspection to picture data D1.By carrying out Fuzzy Processing, launch the reference data D0 of the view data that obtains as RIP, be corrected into for the fog-level of taking the inspection object data D1 same degree that obtains by camera.
Then, rock 132 pairs of inspections of handling part data object D1 execution and rock processing, 133 pairs in comparison process portion has carried out rocking the inspection object data D1 and the reference data D0 execution comparison process of processing.Thus, on the basis of having eliminated the pixel deviation, detect non-accidental gray scale difference.
Then, isolated point is removed handling part 134 and is carried out isolated point removal processing, removes the unwanted isolated point that produces in the comparison process result.Thus, the little isolated point that can avoid should not detecting for the degree of defective is also detected as defective.
Referring again to Fig. 2.Result images generating unit 135 generates result data D2.Result data D2 is such view data, promptly, with with non-defect part (promptly, part beyond the part that extracts as defective) different display mode, defect part among the demonstration inspection object data D1 (promptly highlightedly, to reference data D0 and check that object data D1 has carried out the result of defects detection algorithm, the part that in checking object data D1, is extracted as defective).
In Fig. 4 (a), illustration is by carrying out the result data D2 that the defects detection algorithm obtains to illustrative reference data D0 of Fig. 3 and inspection object data D1.Result data D2, for non-defect part, promptly in checking data object D1, be judged as not and reference data D0 between produce non-accidental difference zone (appropriate area S1, S2, S3), desalination shows, for defect part, promptly be judged as and reference data D0 between produce non-accidental difference zone (defect area T1, T2), by adding the bulk processing overstriking, and with color (for example (peony) demonstration of regulation.
Target data generating unit 140, in the revision directive that receives from the operator result data D2, generation is reflected in view data (hereinafter referred to as " target data D3 ") among the result data D2 with this command content.
In Fig. 4 (b), illustration receives the target data D3 that the operator generates the revision directive of illustrative result data D2 among Fig. 4 (a).For example, in the defect area T1, the T2 that receive from the operator in the result data D2 shown in Fig. 4 (a), defect area T2 is modified under the situation of instruction of appropriate area, target data generating unit 140 generates the target data D3 that defect area T2 is modified to appropriate area shown in Fig. 4 (b).More particularly, with the part of the defect area T2 among the result data D2, be modified to the state that desalination shows, as appropriate area.In addition, the operator can provide the revision directive (with reference to Fig. 7) to result data D2 according to the picture of display result data D2.About this point, describe in more detail in the back.
Adjust image parameter specifying part 150 from the parameter P1～P4 relevant, specify as the parameter more than 1 or 1 (hereinafter referred to as " adjustment image parameter ") of adjusting object with the defect processing Flame Image Process.Adjust the appointment of image parameter, more particularly, undertaken by receive the selection input of wishing the parameter of adjustment parameter P1～P4 from the operator.
Adjust constant determination portion 160 and determine respectively to adjust " adjusting range " and " unit change amount " of image parameter.Determining of adjusting range and unit change amount more particularly, is that the input of the value by receiving adjusting range and unit change amount from the operator is carried out.Here, so-called " adjusting range " is meant the setting range (that is, the scope of the parameter value that parameter adjustment handling part 170 described later makes it to change) of adjusting image parameter, stipulates by the minimum value and the maximal value of parameter value.In addition, so-called " unit change amount " is meant the amplitude of variation (that is the variable quantity when, parameter adjustment handling part 170 makes parameter value variation) of parameter value.
For example, adjusting range at fuzzy parameter P1 is set at " maximal value=8pix " " minimum value=1pix ", the unit change amount is set under the situation of " 1pix " (with reference to Fig. 9), the value that parameter adjustment handling part 170 makes fuzzy parameter P1 from " 1pix " to " 8pix ", with " 1pix " is that unit changes, and determines the value of best fuzzy parameter P1 therein.And, below, will be by the setting range of adjusting range regulation, available parameter value is called " tentative calculation value " when changing with the fixed amplitude of variation of unit change gauge.That is to say that the adjusted value of parameter is defined as some tentative calculation values.
In addition, be under the situation of minimum value setting " 0 ", the adjusted value of parameter also can be defined as " 0 ".Under this situation, parameter value is set at the adjustment of " 0 " and handles, and in the defects detection algorithm, does not in fact carry out.
Parameter adjustment handling part 170 determines to adjust the adjusted value of image parameter.Whether suitably parameter adjustment handling part 170 has parameter value judging part 171 and parameter value determination portion 172.
Whether suitably parameter value judging part 171 changes (more particularly parameter value successively, make parameter value in setting range by the adjusting range regulation, to change (in other words by the fixed amplitude of variation of unit change gauge, the parameter value group (P1, P2, P3, P4) that constitutes in each value by parameter P1～P4 relevant with the defects detection Flame Image Process (promptly, the combination of each value of parameter P1～P4) in, adjustment image parameter value is changed successively)), judge whether the result data D2 that obtains is consistent with target data D3 under each parameter value.That is to say,, judge whether each combination is the combination that can access the parameter value of the result data consistent with target data D3 for all combinations of each tentative calculation value of parameter P1～P4.Wherein, the judgement whether target data D3 and result data D2 be consistent, more particularly, whether consistent the carrying out of the defect part by judging two set of image data.In addition, with obtain by parameter value whether suitably judging part 171 be judged as the parameter value of the result data D2 consistent with target data D3, hereinafter referred to as " appropriate value ".In addition, below can access the parameter value group of the result data D2 consistent, be called " suitable parameter value group " with target data D3.That is to say that the parameter value that constitutes suitable parameter value group is " appropriate value ".For example, in parameter P1～P4, adjusting image parameter is under the situation of parameter P3, is under the parameter value group of " n " in the value of adjusting image parameter P3, carries out the defects detection algorithm.In the defect area part of the result data D2 that obtains thus, under the situation consistent with the defect area part of target data D3, this parameter value group can be described as suitable parameter value group.In addition, for parameter P3, parameter value " n " can be described as appropriate value.
Parameter value determination portion 172 is determined the value (that is adjusted value) as the adjustment result output of adjusting image parameter.More particularly, determine the median of the set that constitutes by suitable parameter value group.That is to say that the median of the set that the appropriate value of selection for adjusting image parameter for each constitutes is defined as adjusted value.For example, be under a kind the situation adjusting image parameter, shown in Figure 11 (a), select tentative calculation value Ta0 as the median of the set Q1 that constitutes by appropriate value Ta, be defined as adjusted value.
＜2. handle
Below, describe for the processing of the parameter adjustment in the defect detecting device 1.
The overall flow that＜parameter adjustment is handled 〉
Fig. 5 is the figure that the overall flow of (adjustment that more particularly, is parameter P1～P4 of relating to of defects detection Flame Image Process is handled) is handled in the parameter adjustment of expression in the defect detecting device 1.
At first, reference data acquisition unit 110 and inspection object data acquisition unit 120 are obtained reference data D0 (with reference to Fig. 3 (a)) respectively and are checked object data D1 (with reference to Fig. 3 (b)), store into (step S1) in the storage part 12.
Then, defects detection handling part 130, reference data D0 and inspection object data D1 to obtaining in step S1 carry out the defects detection algorithm, detect the defective of checking among the object data D1, generate the result data D2 (with reference to Fig. 4 (a)) (step S2) of reflection testing result.
Then, target data generating unit 140 receives revision directive to result data D2 from the operator, generates command content is reflected in target data D3 (with reference to Fig. 4 (b)) (step S3) among the result data D2.
Then, adjust image parameter specifying part 150, from parameter P1～P4 that the defects detection Flame Image Process relates to, specify and adjust image parameter, and determine the adjusting range and the unit change amount (step S4) of each adjustment object.
Then, parameter adjustment handling part 170 determines to adjust the adjusted value (step S5) of image parameter.
Then, parameter adjustment handling part 170, the adjusted value that will determine in step S5, the target data D3 that generates in step S3 and the result data D2 that obtains when carrying out the defects detection algorithm under the parameter value of determining are presented at (step S6) in the display part 14.
Operator from the display result of seeing step S6, reception will will be adjusted under the situation of instruction of adjusted value that the image parameter value be set at demonstration (among the step S7 for being), parameter adjustment handling part 170 is set at the value of adjusting image parameter the adjusted value of determining (step S8) in step S5.More than finishing the adjustment of parameter value handles.Below, more specifically describe for respectively handling of step S3, S4, S5.
The generation of＜2-1. target data D3 is handled 〉
Processing for step S3 more specifically describes.Fig. 6 is the figure of flow process of the processing of expression step S3.
At first, target data generating unit 140 shows the reception picture (step S11) to the revision directive of result data D2 in display part 14.Fig. 7 is the figure that expression receives the configuration example of picture G.Receiving picture G is made of following part: regional g1, and it shows the integral body as the result data D2 that revises object; Zone g2, it amplifies among the display result data D2, as the zone of revising object; And regional g3, it shows the Selection Floater of determining to revise content.
At first, target data generating unit 140, the appointment (step S12) in the zone of object is revised in conduct among the g1 of receiving area.That is to say, the operator observes the integral body of result displayed data D2 in regional g1, whether has the part of wishing correction among the judged result data D2, under the situation of seeing the part of wishing correction, by the drag operation of mouse, specify as the zone of revising object.Target data generating unit 140 receives this zone and specifies.And the zone is specified and also can be used the various sensing equipments (for example trace ball, touch-screen, graphic tablet etc.) except mouse to carry out.
If receive the appointment in the zone of conduct correction object in regional g1, then in regional g2, amplify the zone (step S13) that shows this appointment after the target data generating unit 140.For example, as shown in Figure 7, in regional g1 among the result displayed data D2 under the situation of appointed area U0, U0 inside, viewing area is amplified in the appointment of target data generating unit 140 receiving area U0 in regional g2 the operator.
Then, target data generating unit 140 is received in the appointment (step S14) in the zone (hereinafter referred to as " revising subject area U ") of carrying out revision directive among the regional g2.That is to say that the drag operation by mouse etc. make the operator wish the part of revising more accurately among the designated result data D2 in regional g2, target data generating unit 140 receives the zone of these appointments as revising subject area U.Because in regional g2, amplify the subregion of display result data D2, so the operator can carry out specifying more accurate zone to specify than the zone in regional g1.
Specify if receive the zone of modification region U, then receive the input (step S15) of the correction content that correction subject area U is carried out after the target data generating unit 140.This input operation more particularly is performed such, that is, the Selection Floater of operator to showing among the regional g3, selection and click are used for importing the panel of revising content and " append " the some of " deletion ".
Revise the content input if receive, then target data generating unit 140 is carried out the correction (step S16) to the content of result data D2 indication.For example, select the operator under the situation of panel of " deletion ", target data generating unit 140 will be included in to be revised the interior defect area of subject area U and is modified to appropriate area.Select the operator under the situation of panel of " appending ", target data generating unit 140 will be included in the appropriate area of revising in the subject area data U, be modified to defect area.Below, will in revise subject area, become the actual zone that changes to the object of defect area or appropriate area and be called " difference zone V ".
In addition, in foregoing, revise the appointment (step S14) of subject area U among the temporary transient receiving area g2, the difference zone V that wherein contains is carried out revise (step S16), but also can in regional g2, directly receive the appointment of difference zone V.Under this situation, the operator can provide to give an order, and, for example with the appropriate area that shows among the regional g2 or the part of defect area, is modified to the instruction (for example, with reference to Figure 13 revision directive E3) of defect area or appropriate area that is.
Then, target data generating unit 140 is waited for and is revised the instruction input of finishing (step S18).This instructs input operation, more particularly, undertaken by the panel " in detail " in the Selection Floater that shows among operator's selection and the click on area g3, utilize and be somebody's turn to do " in detail " panel, input expression finishes revision directive, and jumps to and be used for the image instruction that the corresponding parameter of adjusting is set in detail.The image that shows among operator's viewing area g1 judges whether also to exist the part of wishing correction, under the situation that finds the part of also wishing correction, does not select " in detail " panel, but reassigns among the regional g1 as the zone of revising object.Under this situation, target data generating unit 140 is carried out the processing of step S12～S17 again, receives the revision directive to result data D2.On the other hand, finished under the situation that the instruction of all correction position is imported, selected " in detail " panel, the instruction that the input revision directive finishes in operator's judgement.
If receive the input (being) that revision directive finishes in step S18, then target data generating unit 140 is finished the generation of target data D3, and the target data D3 that generates is stored in (step S19) in the storage part 12.More than, the processing of completing steps S3.
＜2-2. adjusts the designated treatment of image parameter 〉
Processing for step S4 more specifically describes.Fig. 8 is the figure of flow process of the processing of expression step S4.
If the processing of completing steps S3 (that is, if in receiving picture G, receiving the operation of " in detail " panel) is then adjusted image parameter specifying part 150 and show the reception picture H (step S21) that the setting of parameter adjustment is imported in display part 14.Fig. 9 is the figure that expression receives the configuration example of picture H.Receive picture H and constitute by following part, that is: regional h1, it receives selection input of adjusting image parameter; Zone h2, it receives the input of adjusting range and unit change amount; And regional h3, the Selection Floater of determining, cancelling of its display setting input.
Adjust image parameter specifying part 150, at first be received in the selection (step S22) that becomes the parameter of adjusting object among the regional h1.More particularly, the display defect detected image is handled the kind of the parameter P1～P4 that relates in regional h1, shows the check box to each parameter P1～P4 simultaneously.The operator utilizes mouse action, chooses by inserting in check box, can need select the parameter of adjustment from parameter P1～P4.
In the example of Fig. 9, in regional h1, show following option: the option of expression fuzzy parameter P1 " blur radius "; The option of parameter P2 " rocking radius " is rocked in expression; And the expression isolated point is removed the option of parameter P4 " isolated point removal pixel count ".In addition, in Fig. 9, though be in hidden state, if the scroll bar that the right side part of drag area h1 shows then can also show the option of expression comparative parameter P3 " tonal range ".The operator can select to wish the parameter of adjustment by choosing desired parameter check box.
Choose some check boxes and selected under the situation of parameter the operator, adjust image parameter specifying part 150, in regional h2, receive input (step S23) for the adjusting range and the unit change amount of the parameter of in step S22, selecting.More particularly, in regional h2, show the input frame of adjusting range (being the maximal value and the minimum value of setting range) and the input frame of unit change amount.The operator can import the adjusting range and the unit change amount of hope by being worth each input frame of input.
For example, as shown in Figure 9, choose the option of " blur radius " the operator, select fuzzy parameter P1 under the situation of adjusting image parameter, adjust image parameter specifying part 150, in the display part of bright apparent " blur radius ", in regional h2, the input frame of the input frame of the adjusting range of display parameter P1 and unit change amount.In Fig. 9, expression input maximal value " 8 (pix) ", minimum value " 1 (pix) " are as the adjusting range of parameter P1, input " 1 (the pix) " state as the unit change amount.
Then, adjust image parameter specifying part 150, wait image parameter to be adjusted to select and adjust the instruction input (step S24) that the setting input of the adjusting range of image parameter and unit change amount is finished about each.This instructs input operation, more particularly, in the Selection Floater of selecting to show among the click on area h3 by the operator, the input expression finishes about the panel of the instruction of the detailed setting of parameter and " determines " to carry out.Also be hopeful adjusting under the situation of the parameter of appending in the object in operator's judgement, do not select " determining " panel, but in regional h1, continue to select parameter.Under this situation, adjust image parameter specifying part 150 and carry out the processing of step S22～S24 again, receive detailed setting input about parameter.On the other hand, under the detailed setting completed situation of operator's judgement about parameter, select " determining " panel, the instruction of end is set in input in detail.
If receive the instruction input (among the step S24 being) that detailed setting finishes, then adjust image parameter specifying part 150, read in the parameter chosen in the check box as adjusting image parameter (step S25), and read in adjusting range and the unit change amount (step S26) of adjusting parameter setting for each.
In addition, if the cancellation panel that shows among the h3 of operator's clicking operation zone then can re-enter parameter selection or adjusting range and unit change amount.Under the situation of operation cancellation panel, the parameter operation of adjusting before image parameter specifying part 150 makes it is invalid.Thus, the processing of end step S4.
Definite processing of＜2-3. adjusted value 〉
Processing for step S5 describes particularly.Figure 10 is the figure of the treatment scheme of expression step S5.
If the processing of completing steps S4 (that is, if in receiving picture H, receiving the operation of " determining " panel) then shows to receive picture G again on display part 14.In receiving picture G,, then begin the processing of step S5 if receive the operation of " rerunning " panel.
At first, parameter adjustment handling part 170 is set the initial value (step S31) of parameter P1～P4 that the defects detection Flame Image Process relate to.More particularly, with the value of the adjustment image parameter among parameter P1～P4, be set at the minimum value of the adjusting range that this adjustment image parameter is set.In addition, for the parameter of non-adjustment image parameter, the value of current setting is set at initial value.The parameter value of non-adjustment object is fixed as this initial value.Below, the parameter value group that will be made of each initial value of parameter P1～P4 is called " initial parameter value group ".
Then, defects detection handling part 130 with the parameter value set (promptly, the initial parameter value group) as judging image parameter value group, judge under the image parameter value group at this, to the reference data D0 and inspection object data D1 that in step S1, obtains, carry out the defects detection algorithm, generate the result data D2 (step S32) of reflection testing result.
Then, the result data D2 that will generate in step S32 is stored in (step S33) in the storage part 12.
Then, whether parameter value suitably judging part 171, and whether relatively result data D2 that generates in step S32 and the target data D3 that generates in step S3, the defect area part of judged result data D2 and the defect part of target data D3 consistent (step S34).More particularly, the part that becomes defect area in result data D2 is defect area in target data D3, in target data D3, do not have simultaneously under other the situation of defect area (promptly, under the on all four situation of defect area that in result data D2 and target data D3, comprises), judged result data D2 is consistent with target data D3, in other cases, judge that two set of image data are inconsistent.
In step S34, judge under the consistent situation of the defect area part of two set of image data, suitably whether parameter value judging part 171, (adjusting image parameter is under a plurality of situation to the parameter value that adopts in the defects detection algorithm that judgement is carried out in step S32, combination for the parameter value that in this defects detection algorithm, adopts) is appropriate value, this parameter value (adjusting image parameter is under a plurality of situation, is the combination of this parameter value) is stored in (step S35) in the storage part 12.In other words, the judgement image parameter value group that adopts in the defects detection algorithm that judgement is carried out in step S32 be suitable parameter value group (promptly, judge that each value that constitutes the adjustment image parameter of judging image parameter value group is appropriate value), each value that is judged as the adjustment image parameter of appropriate value is stored in the storage part 12.
On the other hand, in step S34, judge under the inconsistent situation of defect part of two set of image data, suitably whether parameter value judging part 171, the parameter value that adopts in the defects detection operation rule that judgement is carried out in step S32 (adjusting image parameter is under a plurality of situation, is the combination of the parameter value that adopts in this defects detection operation rule) is not appropriate value (step S36).In other words, judge that the judgement image parameter value group that adopts in the defects detection algorithm of carrying out is not suitable parameter value group in step S32.
Whether then, judge adjusting the value (in other words, in judging image parameter value group, respectively to adjusting the value that image parameter is set) that image parameter is set respectively, be the maximal value (step S37) of the adjusting range of this parameter.
In step S37, the peaked adjustment image parameter that is judged as the adjusting range that is not set at this parameter has at least under the situation more than 1 or 1, carries out the setting change (step S38) of parameter value (that is, constituting the parameter value of judging image parameter value group).More particularly, this processing is following carries out.
Adjusting image parameter is under a kind the situation, with current this is adjusted the value that image parameter is set, and sets the value that changes to the unit change amount size that increases this adjustment image parameter, gets it as new judgement image parameter value group.
But, if for adjusting the value that image parameter is set, be set at the maximal value of the adjusting range of this parameter, then in the moment of finishing the defects detection algorithm under this parameter value, whether suitable judgement finish (promptly for the judgement of all tentative calculation values, in step S37 be), do not carry out its above parameter setting change, but jump to the processing of step S39.
Adjusting image parameter is under two kinds the situation, a parameter value (adjusting image parameter as the 2nd) of adjusting in the image parameter is fixing, with current another is adjusted the value that image parameter (adjusting image parameter as the 1st) is set, setting changes to the value of the unit change amount size that increases this adjustment image parameter, gets it as new judgement image parameter value group.
Adjusting the value that image parameter is set to the 1st, for this is adjusted under the peaked situation of adjusting range of image parameter, adjust the value that image parameter is set with current to the 2nd, setting changes to the value of the unit change amount size that increases this adjustment image parameter, adjust the image parameter value with the 1st simultaneously, be reset to the minimum value of the adjusting range of this adjustment image parameter, get it as new judgement image parameter value group.
But, if value to the 1st, the 2nd each adjustment image parameter setting, all be set at the maximal value of the adjusting range of this parameter, then in the defects detection algorithm moment of finishing under this parameter value, whether suitable judgement is to adjusting all combinations of image parameter tentative calculation value, all finish judgement (that is, in step S37 for being), do not carry out its above parameter setting change, but jump to the processing of step S39.
Adjust image parameter more than or equal to 3 kinds situation under, also with adjust image parameter be two kinds situation in the same manner, set the change parameter value successively, until all combinations, carry out whether suitable judgement to the tentative calculation value of adjusting image parameter.
Referring again to Figure 10.In step S37, judgement is adjusted under the peaked situation of adjusting range that value that image parameters set is this parameter (promptly all, all combinations (all parameter value groups that promptly can obtain) for the tentative calculation value of adjusting image parameter, all carried out under the situation of defects detection algorithm), parameter value determination portion 172 determines to adjust the adjusted value of image parameter then.More particularly, this processing is following carries out.
At first, judge whether to be judged as the parameter value (have under a plurality of situations adjusting image parameter, judge whether to be judged as the combination of the parameter value of appropriate value) (step S39) of appropriate value.More particularly, judge whether store the parameter value that has been judged as appropriate value in the storage part 12.
Under the parameter value that is judged as appropriate value was not stored in situation in the storage part 12, suitably whether parameter value judging part 171, will not find out the advisory of suitable parameter value and gave operator (step S40).For example, the information with " parameter adjustment failure " is presented on the display part 14.
Under the parameter value that is judged as appropriate value is stored in situation in the storage part 12, whether suitably parameter value judging part 171 is chosen as adjusted value (step S41) with the optimum parameter value in the appropriate value (have under a plurality of situations at the adjustment image parameter, be the combination of optimum parameter value).More particularly, the median of selecting to be judged as the tentative calculation value of appropriate value is an adjusted value.
This processing is carried out particularly in the following manner.Adjusting image parameter is under a kind the situation, shown in Figure 11 (a), adjusts among the tentative calculation value T of image parameter at this, selects the median Ta0 of the set Q1 that is made of the tentative calculation value Ta that is judged as appropriate value, is defined as adjusted value.
Adjusting object is under two kinds the situation, shown in Figure 11 (b), adjust image parameter and the 2nd the 1st and adjust among the combination T of tentative calculation value of image parameter, the median Ta0 of the set Q2 that selection is made of the combination Ta that is judged as suitable parameter value is defined as adjusted value.
Adjust image parameter more than or equal to 3 kinds situation under, also with adjust image parameter be two kinds situation similarly, in the combination of the tentative calculation value of adjusting image parameter, select median by the set that constitutes of the parameter value that is judged as appropriate value, be defined as parameter value.More than, the processing of completing steps S5.
＜3. effect 〉
In the flaw detection apparatus 1 that the 1st embodiment of the present invention relates to, when carrying out the adjustment of the parameter value relevant with Flame Image Process, target data generating unit 140 receives revision directive to result data D2 from the operator, generate target data D3, the parameter value that parameter adjustment handling part 170 can obtain the result data D2 consistent with target data D3 is defined as adjusted value.Thus, even the operator does not directly set input parameter value (that is, even set input without the method for trial and error desired parameter value of searching),, just can easily obtain the parameter value of expecting only by providing revision directive to result data D2.In addition, because parameter adjustment handling part 170 is defined as obtaining the value of the result data D2 consistent with target data D3 with adjusted value, so can carry out parameter adjustment exactly.
In addition, owing to can easily carry out suitable parameter adjustment, current many parameter adjustments of being undertaken by method of trial and error can be changed to user task.Thus, can cut down maintenance cost.
Especially, in flaw detection apparatus 1, because can be easily and adjust the parameter P1～P4 of the detection sensitivity of regulation defective exactly, so in flaw detection apparatus 1, the operator can detect defective with desired detection sensitivity.
Especially, be presented on the display part 14, from display frame, receive revision directive input, so the operator can be easily and imported revision directive to result data D2 exactly result data D2 because target data generating unit 140 will be checked object data D1.Its result can be easily and carry out parameter adjustment exactly.
Especially, because whether suitably parameter value judging part 171 changes the value of adjusting image parameter successively, judge simultaneously parameter value whether be appropriate value (more particularly, by judging the result data D2 that under each parameter value, obtains, whether consistent with target data D3, judge whether this parameter value is appropriate value), the adjusted value that parameter value determination portion 172 will be adjusted image parameter is defined as appropriate value, so can detect the parameter value that can obtain the result data D2 consistent with target data D3 exactly.
Especially, because parameter value determination portion 172, the median of the set that selection is made of the parameter value that is judged as appropriate value, the adjusted value of adjusting image parameter is defined as adjusted value, so, in appropriate value, also can determine optimum value (promptly, obtain the parameter value of the result data D2 consistent the most reliably with target data D3), as the adjusted value of parameter.
Especially, because adjust the input of constant determination portion 160 according to the operator, regulation is adjusted the adjusting range of image parameter, thus can be in scope be adjusted, the parameter value that reduction can be obtained.Thus, can more promptly carry out parameter adjustment.
Especially, because adjust image parameter specifying part 150, according to operator's input, from the parameter that the defects detection Flame Image Process relates to, specify, so can reduce the kind that the parameter of object is adjusted in conduct as the parameter of adjusting object.Thus, can promptly carry out parameter adjustment.
(the 2nd embodiment)
＜1. structure 〉
The flaw detection apparatus of a mode of the data processing equipment that relates to as the 2nd embodiment of the present invention because with the same apparatus structure of the flaw detection apparatus that in the 1st embodiment, illustrated 1 on realize, so omit its explanation.In addition, in following content, when the textural element of the flaw detection apparatus that relates to of expression the 2nd embodiment, for the same textural element of the flaw detection apparatus that relates to embodiment 11, use in the explanation of the 1st embodiment, use with reference to label.
The structure that＜1-1. is relevant with parameter adjustment 〉
The flaw detection apparatus that present embodiment relates to has the function of the parameter of adjusting the Flame Image Process that the inspection of inspection version/printed matter relates to.Figure 12 is the explanation structure relevant with the parameter adjustment function.
The defect detecting device that present embodiment relates to, defect detecting device 1 difference that relates to the 1st embodiment are, replace and adjust image parameter specifying part 150, have the image parameter of adjustment judging part 250.In the content below, the difference of the flaw detection apparatus 1 that relates to the 1st embodiment only is described, omits explanation for identical point.In addition, for same textural element, the label that mark is identical is represented.
Adjust image parameter judging part 250 and adjust image parameter specifying part 150 similarly, from parameter P1～P4 that the defects detection Flame Image Process relates to, specify and adjust image parameter.But adjusting image parameter judging part 250 is not to specify by the selection input that receives the operator to adjust image parameter, but according to the operator that receives in target data generating unit 140 revision directive to result data D2, specifies the adjustment image parameter.
Adjusting image parameter judging part 250 has: characteristic quantity acquisition unit 251, relevant parameter judging part 252, adjustment image parameter determination portion 253.
Characteristic quantity acquisition unit 251 according to the operator that the receives revision directive (being designated hereinafter simply as " revision directive ") to result data D2, is obtained the characteristic quantity of regulation respectively in target data generating unit 140.
Here, 13～Figure 15 more specifically describes characteristic quantity with reference to the accompanying drawings.Illustration result data D2 in Figure 13 (a) wherein, illustration target data D3 in Figure 13 (b), this target data D3 receive the operator revision directive E1～E3 of the illustrative result data D2 of Fig. 3 (a) are generated.
Here, revision directive E1 is 1 isolated point (difference zone V1) of revising in the subject area U1 being in, is modified to the instruction of defect area from appropriate area.In addition, revision directive E2 is a plurality of isolated points (difference zone V2a, V2b, V2c, V2d) of revising in the subject area U2 being in, is adapted to the instruction of defect area from appropriate area.In addition, revision directive E3 is the instruction that difference zone V3 is adapted to defect area from appropriate area.
The characteristic quantity of so-called revision directive is meant following value,, in order to generate the result data D2 corresponding with this revision directive content, and is used to specify the judgement key element of the parameter kind that should adjust that is.Characteristic quantity acquisition unit 251 is obtained area S, number A, these 3 information of Luminance Distribution H as characteristic quantity from each revision directive.
Area S is the total area of the difference zone V that relates to of 1 revision directive.For example, by the area S that revision directive E2 obtains, be the summation of revising the area of each difference zone V2a～V2d that contains among the subject area U2.In addition, for example,, be the area of difference zone V3 according to the area S that revision directive E3 obtains.
Number A is the number of the difference zone V that relates to of 1 revision directive.For example, as revision directive E2, exist in revising subject area U2 under the situation of 4 difference zone V1a～V1d independently, number A is " 4 ".On the other hand, shown in revision directive E1, only exist in revising subject area U1 under the situation of 1 difference zone V1 independently, number A is " 1 ".In addition, shown in revision directive E3, under the situation of directly specifying difference zone V, number A is " 1 ".
Luminance Distribution H is the Luminance Distribution of the regulation zone of checking object data D1 (that is, being meant that center from difference zone V is to the zone that occupies specialized range, hereinafter referred to as " amount of brightness is obtained regional W ").
Here, describe for obtaining more specifically of Luminance Distribution H.When obtaining Luminance Distribution H, adjustment image parameter judging part 250 at first given luminance amount obtains regional W.More particularly, at first, specify the center Vo (under the situation that has a plurality of difference zone Vs relevant, being the center of the regional V of all differences) of difference zone V, stipulate that center Vo with appointment is that the amount of brightness at center is obtained regional W with 1 revision directive.
Then, adjust image parameter judging part 250 and obtain the Luminance Distribution that the amount of brightness of checking among the object data D1 is obtained regional W.For example, under the situation of revision directive E1, amount of brightness is obtained the center of regional W1 and the center Vo consistent (with reference to Figure 14 (a)) of difference zone V1.,, then obtain the Luminance Distribution of regional W here, can obtain for example histogram shown in Figure 15 (a) as amount of brightness if in checking object data D1, shown in Figure 14 (a), in the background of white, there are differences the point of regional V1.
In addition, for example under the situation of revision directive E3, amount of brightness is obtained the center of regional W and the center Vo consistent (with reference to Figure 14 (b)) of difference zone V3., checking among the object data D1 that be under the situation of boundary part of pattern and background, the lightness distribution of then obtaining regional W as the lightness amount can obtain for example histogram shown in Figure 15 (b) here shown in Figure 14 (b) at difference zone V3.
On the other hand, in checking object data D1, difference zone V3 is under the situation of the area part that gradually changes shown in Figure 14 (c), obtains the Luminance Distribution of regional W as amount of brightness, can obtain for example histogram shown in Figure 15 (c).
Referring again to Figure 12.Relevant parameter judging part 252 is according to each characteristic quantity that is obtained by characteristic quantity acquisition unit 251, the kind of the parameter that judgement should be adjusted.
Here, more specifically describe for determination methods.Carry out based on the judgement of area S is following.At first, judge that whether the value of area S is greater than setting s1.Here, under the situation of value greater than setting s1 of judging area S, judgement should be adjusted comparative parameter P3.
Be not more than under the situation of setting s1 in the value of judging area S, afterwards, whether the value of judging area S is greater than setting s2 (wherein, s1＞s2).Here, under the situation of value greater than setting s2 of judging area S, judgement should be adjusted fuzzy parameter P1.On the other hand, be not more than under the situation of setting s2 in the value of judging area S, judgement should be adjusted isolated point and remove parameter P4.
Carry out based on the judgement of number A is following.At first, judge that whether the value of number A is greater than setting n1.Here, under the situation of number A greater than setting n1, judgement should be adjusted and rock parameter P2.On the other hand, judging that number A is not more than under the situation of setting n1, be judged as and adjust isolated point removal parameter P4.
Carry out based on the judgement of Luminance Distribution H is following.At first, adjust image parameter judging part 250, judge the histogram shape of the monochrome information of obtaining, be among " the 1st distribution shape h1 " " the 2nd distribution shape h2 " " the 3rd distribution shape h3 " which.
Consequently, judging that histogram shape is under the situation of the 1st distribution shape h1, judgement should be adjusted isolated point and remove parameter P4, judging that histogram shape is under the situation of the 2nd distribution shape h2, judgement should be adjusted fuzzy parameter P1, judging that histogram shape is under the situation of the 3rd distribution shape h3, judgement should be adjusted comparative parameter P3.
Here, for judging that histogram shape is the method for any distribution shape h1～h3, describe with reference to Figure 15, Figure 16.Wherein, Figure 16 is the figure that the flow process of histogram shape kind is judged in expression.
At first, in histogram distribution, obtain maximum brightness value (maximal value M) (step S101).
Then, calculate reference value N (step S102).Wherein, reference value N is half the value (N=M/2) of maximal value M.
Then, judge that brightness is the dispersion of distribution of the position of reference value N, whether greater than setting (step S103).Wherein, shown in Figure 15 (b), be under the situation of position appearance more than or equal to 2 independent distribution zones of reference value at brightness value, judge that whether the wideest dispersion of distribution is greater than setting.
In step S103, judge (that is, under the situation of existence greater than the dispersion of distribution of Rack) under the situation of the dispersion of distribution greater than setting, judge that histogram distribution is the 3rd distribution shape h3 (step S104).
In step S103, judge that the dispersion of distribution is not more than under the situation of setting (that is, not existing under the situation greater than the dispersion of distribution of Rack), judges whether to occur more than or equal to 2 distributed areas (step S105) afterwards.
In step S105, judge under the situation that the distributed areas more than 2 or 2 occur, judge that histogram distribution is the 2nd distribution shape h2 (step S106).
In step S105, judge not occur under the situation of the distributed areas more than 2 or 2, judge that histogram distribution is the 1st distribution shape h1 (step S107).
For example, the histogram shape of Figure 15 (a)～(c) is being carried out under the situation of above-mentioned judgment processing, judging that the histogram of Figure 15 (a) is the 1st distribution shape h1, should adjust isolated point and remove parameter P4.In addition, judge that the histogram of Figure 15 (b) is the 2nd distribution shape h2, should adjust fuzzy parameter P1.In addition, judge that the histogram of Figure 15 (c) is the 3rd distribution shape h3, should adjust comparative parameter P3.
Referring again to Figure 12.Adjust image parameter determination portion 253, the judged result of comprehensive relevant parameter judging part 252 outputs, the final appointment adjusted image parameter.More particularly, calculate the shared ratio of each parameter in the judged result of storage, proportionally Gao Di order is defined as adjusting image parameter with 2 parameters.
＜2. handle
Parameter adjustment in the flaw detection apparatus 1 that relates to for the 2nd embodiment is handled and is described.The whole flow process that parameter adjustment is handled as shown in Figure 5.Wherein, in the present embodiment, adjust following the carrying out of designated treatment (that is, being equivalent to the processing of step S4) of image parameter.
＜2-1. adjusts the designated treatment of image parameter 〉
Figure 17 is the flow process that is illustrated in the designated treatment of the adjustment image parameter that present embodiment carries out.
At first, characteristic quantity acquisition unit 251 is obtained the characteristic quantity (that is, area S, number A, Luminance Distribution H) (step S111) of regulation from each revision directive to result data D2 that is received by target data generating unit 140 among step S3.
For example, when the generation of target data D3, as shown in figure 13, receive under the situation of 3 revision directive E1～E3, characteristic quantity acquisition unit 251 for each revision directive E1～E3, is obtained area S, number A, Luminance Distribution H respectively.
Then, corresponding parameter judging part 252 is according to each characteristic quantity that obtains in step S111, the parameter kind (step S112) that judgement should be adjusted.And judged result is sent to from corresponding parameter judging part 252 in order and adjusts image parameter determination portion 253, adjusts image parameter determination portion 253 storage judged results.
If finish the judgement of all characteristic quantities that corresponding parameter judging part 252 is obtained, then judge to adjust image parameter determination portion 253 and obtained judged result to all characteristic quantities (at step S113 for being), adjust the judged result of image parameter determination portion 253 comprehensive storages, specifying becomes the parameter (step S114) of adjusting object.More particularly, calculate in the judged result of storage, the ratio that each parameter is shared, proportionally Gao Di order is defined as adjusting image parameter with 2 parameters.。
For example, in step S111, in characteristic quantity acquisition unit 251 respectively to revision directive E1～E3, obtain under the situation of area S, number A, Luminance Distribution H, in step S112, corresponding parameter judging part 252 is judged the kind of the parameter that should adjust successively respectively according to 9 characteristic quantities that obtain.This judged result sends the adjustment image parameter to and determines 253, adjusts image parameter determination portion 253 and stores 9 judged results as shown in figure 18.In the example of Figure 18, the parameter that will account for maximum ratio in the judged result of storage is that isolated point is removed parameter P4 and second largest fuzzy parameter P1 of ratio, is defined as adjusting image parameter.
Referring again to Figure 17,, then adjust image parameter judging part 250 and on display part 14, show the reception picture H (with reference to Fig. 9) (step S115) that the setting of parameter adjustment is imported if determined that in step S114 2 are adjusted image parameter.And, in the present embodiment, receive the regional h1 among the picture H, not the zone that is used to receive the selection input of adjusting image parameter, but being used to show the definite result's who adjusts image parameter zone, conduct works.That is to say, at this moment, adjust the reception picture H that image parameter judging part 250 shows following state here, in being shown in the check box of regional h1, selected state with the check box that in step S114, is defined as adjusting the image parameter corresponding parameters.For regional h2, regional h3, identical with the 1st embodiment.
Then, adjust image parameter judging part 250, receive corresponding respectively the adjust adjusting range of image parameter and the input (step S116) of unit change amount.This handles identical with the processing of the step S23 (Fig. 8) that illustrates before.
Then, finish the adjusting range of each adjustment image parameter and the input of unit change amount, receive under the situation of the operation of determining panel (in step S117 for being), adjust image parameter judging part 250 and read in adjusting range and the unit change amount (step S118) of each being adjusted the image parameter setting.Thus, finish the designated treatment of adjusting image parameter.
＜3. effect 〉
In the flaw detection apparatus that the 2nd embodiment of the present invention relates to, especially, because adjust image parameter judging part 250 according to revision directive from the operator, from the parameter that the defect inspection Flame Image Process relates to, determine to become the parameter of adjusting object, so can suitably select the parameter kind that to adjust, to obtain the result data D2 consistent with target data D3.Thus, can be rapidly and carry out parameter adjustment exactly.
(the 3rd embodiment)
＜1. structure 〉
The flaw detection apparatus of a mode of the data processing equipment that relates to as the present invention's the 3rd embodiment because with the same apparatus structure of the flaw detection apparatus that in the 1st embodiment, illustrated 1 in realize, so omit its explanation.In addition, in following content, when the inscape of the flaw detection apparatus that the 3rd embodiment relates to is described, use in the explanation of the 1st embodiment used with reference to label.
The flaw detection apparatus that present embodiment relates to, the flaw detection apparatus that relates to the 1st embodiment similarly have the constant of adjustment determination portion 170.But in the present embodiment, adjustment constant determination portion 170 is not to determine adjusted value by the median of the set of selecting to be made of appropriate value, but the appropriate value that primary detection goes out is defined as adjusted value.
＜2. handle
Parameter adjustment in the defect detecting device 1 that relates to for the 3rd embodiment is handled and is described.The overall flow that parameter adjustment is handled as shown in Figure 5.Wherein, in the present embodiment, definite processing of parameter adjustment value (that is, being equivalent to the processing of step S5) is carried out in the following manner.
Definite processing of＜2-1. adjusted value 〉
Figure 19 is the figure of flow process of definite processing that represents the adjusted value of the adjustment image parameter carried out in the present embodiment.
At first, carry out and the same processing (step S211～step S214) of step S31～S34 (Figure 10) of explanation before.
In step S214, under the situation of the part unanimity of the defect area of judging two view data, parameter value whether suitably judging part 171 judges that the parameter value that adopts in the defects detection algorithm of carrying out (adjusting image parameter is under a plurality of situation, is the combination of the parameter value that adopts) is appropriate value in this defects detection algorithm in step S212.In this case, the parameter value that is judged as this appropriate value is defined as adjusting the adjusted value (step S215) of image parameter.That is to say, be not shown in the 1st embodiment, the defects detection algorithm carried out in each all combinations of adjusting the tentative calculation value of image parameter, and in the moment of finding out suitable parameter value, this parameter value is defined as adjusted value.That is to say, will begin detected appropriate value most and be defined as adjusted value.
On the other hand, in step S214, judge under the inconsistent situation of defect area part of two view data, parameter value whether suitably judging part 171 judges that the parameter value that adopts in the defects detection algorithm of carrying out (adjusting image parameter is under a plurality of situation, is the combination of the parameter value that adopts in this defects detection algorithm) is not appropriate value (step S216) in step S212.
Whether under this situation, judge afterwards for adjusting the value that image parameter is set respectively, be the maximal value (step S217) of the adjusting range of this parameter.
In step S217, judge to have at least under the peaked adjustment image parameter situation of the adjusting range that is not set at this parameter more than 1 or 1, carry out the setting change (step S218) of parameter value.This handles identical with the step S38 (Figure 10) of explanation before.
In step S217, judgement is adjusted under the peaked situation that value that image parameters set is this parameter adjustment amplitude (promptly all, all combinations to the tentative calculation value of adjusting image parameter have been carried out under the situation of defects detection algorithm), whether suitably parameter value judging part 171 advisory that will not find out suitable parameter value gives operator (step S219).That is to say,, also carry out the processing of step S19 even, also do not find out under the situation of suitable parameter value in that the defects detection algorithm has been carried out in each all combinations of adjusting the tentative calculation value of image parameter.Thus, finish to adjust definite processing of adjusted value of image parameter.
＜3. effect 〉
In the flaw detection apparatus that the 3rd embodiment of the present invention relates to,, the parameter value that begins to be judged as appropriate value most is defined as adjusted value, so can carry out parameter adjustment rapidly because for the adjusted value of adjusting image parameter.
(the 4th embodiment)
＜1. structure 〉
The flaw detection apparatus of an embodiment of the data processing equipment that relates to as the 4th embodiment of the present invention because with the same apparatus structure of the flaw detection apparatus that in the 1st embodiment, illustrated 1 on realize, so omit its explanation.In addition, in following content, when the inscape of the flaw detection apparatus that relates to of explanation the 4th embodiment, use use in the explanation of the 1st embodiment with reference to label.
The structure that＜1-1. parameter adjustment is relevant 〉
The flaw detection apparatus that present embodiment relates to has the function of adjusting the Flame Image Process parameter that the inspection of inspection version/printed matter relates to.Figure 20 is the relevant structural drawing of explanation parameter adjustment function.
The flaw detection apparatus that present embodiment relates to, its mode of determining the parameter adjustment value is different with the flaw detection apparatus 1 that the 1st embodiment relates to.In following content, the difference of the flaw detection apparatus 1 that relates to the 1st embodiment only is described, omit explanation for identical point.In addition, for same inscape, mark identical label.
The flaw detection apparatus that present embodiment relates to has the function of the parameter of adjusting the Flame Image Process relevant with examining the inspection of version/printed matter.Figure 20 is the explanation structural drawing relevant with the parameter adjustment function.In addition, in following content, suitably with reference to Figure 21, Figure 22.Figure 21, Figure 22 are that explanation uses the adjusted value of genetic algorithm to determine the concept map of mode.
The flaw detection apparatus that present embodiment relates to is provided with parameter adjustment handling part 470.Parameter adjustment handling part 470 adopts genetic algorithm to determine to adjust the adjusted value of image parameter.More particularly, from the multiple parameter values group corresponding, utilize genetic algorithm to extract the parameter value group that can obtain the result data D2 consistent with target data D3 with all combinations of each tentative calculation value of adjusting image parameter.And, with the value of the formation parameter value group that extracts, be defined as adjusting the adjusted value of image parameter.
Parameter adjustment handling part 470 is provided with as shown in figure 20: the 1st generation group of individuals generating unit 471, matching value acquisition unit 472, pairing generating unit 473, cross processing portion 474, sudden change execution portion 475, optimum parameter value determination portion 476.
The 1st generation group of individuals generating unit 471, (wherein, N is the natural number of regulation, and its value can be set arbitrarily to generate stated number N.In addition, in the example of Figure 21, N=4) group of individuals is as the group of individuals acquisition of the 1st generation.Each individuality has the chromosome that is made of M gene.Wherein, as gene, more particularly, be the some values in 0.0～1.0 scope, the genetic number that each individuality in the 1st generation has is selected respectively at random.In addition, the genetic number M that contains in the chromosome is (in the example of Figure 21, M=7), the number of parameters that relates to the defects detection Flame Image Process is consistent, as shown in figure 22, and each gene 1,2 ... 7, corresponding with some parameters that the defects detection Flame Image Process relates to.That is to say that each chromosome is corresponding with the parameter value group.
In addition, under the situation of the parameter more than 1 or 1 that will from the parameter that the defects detection Flame Image Process relates to, select, make each gene corresponding with some parameter adjustment objects as the adjustment image parameter.That is to say that under this situation, genetic number will be consistent with the number of adjusting image parameter.
Matching value acquisition unit 472 is obtained individual matching value.Matching value acquisition unit 472 by respectively N the individuality that constitutes each generation being carried out processing described later, is obtained N individual matching value separately.Then, the matching value that obtains is individual related with each, be stored in the storage part 12.
The processing of obtaining matching value is described with reference to Figure 22.When obtaining matching value, (in the example of Figure 22, M=7) gene obtain M parameter value to explain the M that contains in the individuality at first respectively.That is to say, obtain 1 parameter value group by 1 chromosome.Explain the processing of the gene and the value of getting parms, the calculation process shown in following by carrying out (formula 1) is carried out.Wherein, in formula (1), " Gi " is genetic value, " Pi " be with the value of this gene corresponding parameters (i=1,2 ..., M).In addition, " min " reaches " max ", is respectively minimum value and maximal value to the adjusting range of this parameter setting.In addition, " div " is the unit change amount to this parameter setting.The calculation process of the following value of radix point is given up in " int{} " expression in addition.
Pi=min+ (max-min) * int{Gi*div}/div ... (formula 1)
Then, under the parameter value group of obtaining, carry out the defects detection algorithm, obtain result data D2.And, result data D2 and the target data D3 that obtains compared, calculate individual matching value.Wherein, so-called be somebody's turn to do " matching value " be meant the area occupation ratio of the part that result data D2 is consistent with destination image data D3 (area of uniform portion is with respect to the ratio of the area of integral image).That is to say, individual matching value, be that expression can obtain and the result data D2 of this individuality (chromosome) corresponding parameters value group and the value of target data D3 consistent degree, as can be known should value big more, good individuality (with the corresponding individuality of parameter value group that can access with the highly consistent result data D2 of target data D3).Wherein, matching value can be got the value of 0～1 scope.For example, we can say that matching value is the individuality of " 1 ", can be described as and the parameter value group the best corresponding individuality that can obtain with the on all four result data D2 of target data D3.
Pairing generating unit 473, by N individuality with generation, (wherein, this is that N is the situation of even number to generate ((N/2)-1) group.At N is under the situation of odd number, is (N-1)/2 group) pairing.More particularly, at first,, from N group of individuals, select (N-2) individual (is under the situation of odd number at N, for (N-1) individual) individuality with generation according to the principle of natural selection that has used matching value.Here selected individuality becomes the female individuality that generates individuality of future generation.So-called principle of natural selection, be meant and make good individuality (promptly, here be the high individuality of matching value) the principle preferentially inherited of gene to the next generation, by making in N individuality with generation, good more individuality is chosen as high more realization of probability of female individuality.For example, with matching value as being chosen as female individual selection probability.In addition, also can adopt following structure, it is according to the matching value size sequence, and tagging successively begins to distribute successively high selection probability from the individuality of height ordering.Pairing generating unit 473 makes mutually that 2 different individualities are a pair of from (N-2) that select according to above-mentioned principle individual (N is individual for (N-1) under the situation of odd number) individuality, it is right to generate ((N/2)-1) group (N is (N-1)/2 under the situation of odd number) assembly.
And, by carrying out intersection described later sudden change, generate follow-on individuality, with the number of individuals of future generation that generates for last generation number of individuals be the mode (that is) of equal number no matter to pass through several generations, the mode that number of individuals is all constant, carry out the number of individuals adjustment.In these cases, because the number of individuals of use (that is, as described later in intersecting, by the number of individuals that intersect to generate), be individual under the situation of even number at N for (N-2), be individual under the situation of odd number at N for (N-1), so directly the number of individuals of future generation of Sheng Chenging will reduce.Therefore, the adjustment that replenishes this deficiency is handled.
At N is under the situation of even number, lacks two individualities.Under this situation, for example, utilize elite (elite) method (the individual method that directly keeps that matching value is high), replenish 1 individuality, replenish 1 simultaneously and have the genetic individuality of selecting at random as individuality of future generation.On the other hand, be under the situation of odd number at N, lack 1 individuality.Under this situation, for example, utilize elite's method to replenish 1 individuality.
And also can make the pairing number that generates herein (is that N is the situation of even number for (N/2) group here.At N is under the situation of odd number, is ((N+1)/2) group).Under this situation, the number of individuals that uses in the intersection because at N be under the situation of even number for N, be under the situation of odd number at N, individual for (N+1), so be under the situation of odd number at N, the number of individuals of future generation of generation increases.Therefore, under this situation, be under the situation of odd number at N, after cross processing, from the individuality of future generation that generates, select 1 individual deletion at random, carry out the number of individuals adjustment.
In addition, as mentioned above,, carry out the number of individuals adjustment here in the follow-on number of individuals mode identical that generates with the number of individuals in last generation.That is to say that the quantity of the individuality that uses in the intersection under the situation that is (N-2), is the individual mode of N with the number of individuals that generates, and utilizes elite's method and gene selection at random, replenishes the individuality of 1 deficiency respectively.In addition, the number of individuals that uses in intersection is under (N-1) individual situation, utilizes elite's method to replenish the individuality of 1 deficiency.And the quantity of the individuality that uses in intersection is under (N+1) individual situation, selects 1 individuality and deletion from the individuality of future generation that generates at random.
Sudden change execution portion 475 intersects when generating new group of individuals utilizing, and in the gene of the new group of individuals of this generation, produces sudden change with the probability of regulation.So-called sudden change is meant the change of newly-generated N individual genetic value.By make genetic gene mutation (that is) with the probability of stipulating with the genetic value of probability randomly changing of regulation, can eliminate genetic deviation, avoid in optimization process, being absorbed in the danger of local minimum (local minimum).
Optimum parameter value determination portion 476, definite adjusted value (as the value of adjusting result's output) of adjusting image parameter.More particularly, if generate regulation certain generation (for example, T generation (wherein, T be set arbitrarily more than or equal to 2 natural number)) individuality, then in contemporary individuality, extraction has the individuality of the highest matching value as optimized individual.Then, judgement is an optimum parameter value set with the chromosome corresponding parameters value group of the optimized individual of extracting, and with constituting M parameter value of this optimum parameter value set, is defined as each parameter adjustment value.In other words, will explain M the parameter value that M of containing in the optimized individual chromosome is genetic and obtain, be defined as each parameter adjustment value.Also can constitute in addition,, have moment of individuality of the matching value of the value that is higher than regulation in generation, extract this individuality as optimized individual with algebraic independence ground.
＜2. handle
Parameter adjustment in the flaw detection apparatus that relates to for the 4th embodiment is handled and is described.The flow process of the integral body that parameter adjustment is handled as shown in Figure 5.Wherein, in the present embodiment, carry out definite processing (that is, being equivalent to the processing of step S5) of adjusted value in the following manner.
Definite processing of＜2-1. adjusted value 〉
Figure 23 is the figure of flow process that represents the predetermined processing of the adjusted value carried out in the present embodiment.And, below also suitably with reference to Figure 21, Figure 22.
If the processing of completing steps S4 (Fig. 5) then shows to receive picture G again in display part 14.Receiving on the picture G,, then begin the processing of step S5 if receive the operation of " rerunning " panel.
At first, the 1st generation group of individuals generating unit 471, generate N individual (in the example of Figure 21, being individuality 1,2,3,4) (step S311) constituting for the 1st generation.Here each individuality of Sheng Chenging as mentioned above, has M the chromosome of selecting by at random that gene constitute.
Then, matching value acquisition unit 472 is obtained N the individual matching value separately that constituted for the 1st generation, and the matching value that obtains and each individuality is related and be stored in (step S312) in the storage part 12.
Then, pairing generating unit 473 is individual by N of the 1st generation, generates the pairing (step S313) that ((N/2)-1) organizes (is under the situation of odd number at N, for (N-1)/2 group).More particularly, according to the principle of natural selection that has used matching value, selecting (N-2) from N the group of individuals in the 1st generation, individual (at N is under the situation of odd number, for (N-1) individual) individuality, individual (at N is under the situation of odd number by (N-2) that select, for (N-1) individual) individuality, make 2 mutually different individual pairings, generate the pairing of ((N/2)-1) group (be under the situation of odd number for (N-1)/2 group) at N.
Then, cross processing portion 474, for each pairing that generates in step S313, making becomes the individual intersection of 2 of pairing, generates new individuality (step S314).More particularly, as shown in figure 21, replace and to compare the gene that are positioned at latter half, regenerate the individuality of (N-2) individual (being individual under the situation of odd number) for (N-1) at N with the crossover location of the regulation of from chromosome, selecting at random.And,,, carry out the number of individuals adjustment here in the follow-on number of individuals mode identical that generates with the former generation number of individuals.That is to say, be under the situation of (N-2) at the number of individuals that generates, and utilizes elite's method, replenishes 1 individuality, replenishes 1 simultaneously and have the genetic individuality of selecting at random.On the other hand, be under (N-1) individual situation at the number of individuals that generates, utilize elite's method to replenish 1 individuality.
Then, sudden change execution portion 475 makes the new individuality that generates in step S314, produce genetic sudden change (step S315) with the probability of stipulating.
Utilize the processing of step S314～step S315, regenerate N individuality, the N of Sheng Chenging individual (in the example of Figure 21, being individual 1 ', 2 ', 3 ', 4 ') is the group of individuals that constituted for the 2nd generation herein.
Then, optimum parameter value determination portion 476 judges whether to generate the individuality (step S316) in certain generation (T generation (wherein, T be the natural number of setting) arbitrarily) of regulation.In other words, judge processing, whether carried out regulation time (T-1 time) from step S312 to step S315.
In step S316, judge under the situation of the individuality do not generate T generation, return the processing of step S312 again.For example, under the situation of T=3,, generate the group of individuals in the 2nd generation, then return the processing of step S312 again if in the processing of step S315.In this case, by to the 2nd generation group of individuals, the processing of execution in step S312～step S315 generate to constitute the group of individuals in the 3rd generation.
In step S316, judge to generate under the situation of individuality in T generation, matching value acquisition unit 472 is obtained and is constituted T N the individual matching value separately in generation, optimum parameter value determination portion 476, T extract in for individuality have the highest matching value individuality as optimized individual (step S317).For example, under the situation of T=3, if in the processing of step S315, generate the 3rd generation group of individuals, then optimum parameter value determination portion 476 in the 3rd generation individuality, extract have the highest matching value individuality as optimized individual.
In addition, also can the judgment processing of step S316 be changed, after the processing of step S312 (that is, after obtaining each individual matching value), carry out the processing that judges whether to exist individuality with the matching value that is higher than setting.In this case, exist in judgement under the situation of this individuality, extract this individuality as optimized individual.
Then, optimum parameter value determination portion 476 is explained M gene that contain in the chromosome of the optimized individual that extracts in step S317, obtain M parameter value, this value of obtaining is defined as the adjusted value (step S318) of each parameter.Thus, the processing of completing steps S5 (Fig. 5).
And in foregoing, follow-on individuality constitutes, and is matched by the individuality of selecting by the principle of natural selection that has used matching value, this pairing is intersected and generate, but the generation method of individuality of future generation is not limited thereto.For example, also can adopt the high individuality of direct reservation matching value as individual method (elite's method) of future generation.In addition, also can adopt the method that generates individuality of future generation at random.
＜3. effect 〉
In the flaw detection apparatus that the 4th embodiment of the present invention relates to, because adopt genetic algorithm to obtain the adjusted value of adjusting image parameter, so can promptly obtain optimum parameter value.
In addition, in the flaw detection apparatus that present embodiment relates to, a plurality of gene that various parameters and chromosome are contained are related respectively (promptly, make that each is individual related with the combination of each parameter value), use genetic algorithm to obtain the high defect individual of matching value (that is, can draw the combination of the best adjusted value of target data D3).Therefore, increase, and can obtain best adjusted value rapidly even adjust the kind of image parameter.
(variation)
＜the 1 variation 〉
In the respective embodiments described above, in the defects detection algorithm, order is carried out Fuzzy Processing, is rocked processings in accordance with regulations, contrast is handled, isolated point is removed and handled, but the defects detection algorithm of execution is not limited thereto in flaw detection apparatus 1.For example, can replace above-mentioned each Flame Image Process, specify Fuzzy Processing or overstriking treatment combination in the defects detection algorithm core, perhaps they and above-mentioned each Flame Image Process are to use simultaneously.And, can utilize aforesaid way, be adjusted at parameters needed in the various processing of combination.That is to say, in the above-described embodiment, a plurality of parameters as the defects detection Flame Image Process relates to have listed fuzzy parameter P1, have rocked parameter P2, comparative parameter P3, isolated point removed parameter P4, but the parameter that the defects detection Flame Image Process relates to is not limited thereto.
For example, can be with following various parameters as adjusting image parameter, that is: the filtration radius value in the filtration treatment, color gamut (for example, the L value in the Lab color space that in the extraction of the color gamut of appointment is handled, should extract, each scope of a value and b value), add the overstriking amplitude in the bulk processing, refinement amplitude in the thinning processing, filtration radius value and filter-id value during the detection of marginal element is handled, whether regulation carries out the mark value of the various processing of synthetic processing of shielding or comparison process etc., areal extent in the area judgment processing (each value of minimum area and maximum area), largest contours length value during Characteristic Extraction is handled etc.
In addition, just defects detection is not handled the parameter that relates to, also can with obtain when checking object data D1 various parameters (for example, the light angle of image analyzer 5, light source colour, optics multiplying power, image reads or the processing of noise remove etc. relates to various adjustment variablees etc.), be included in and adjust in the image parameter.
In addition, also can comprise and whether carry out defects detection algorithm process (opening/closing of algorithm process function itself), as adjusting one of image parameter.
＜the 2 variation 〉
In the respective embodiments described above, for in flaw detection apparatus 1 as an embodiment of data processing equipment, the situation of carrying out the parameter adjustment that the defects detection Flame Image Process relates to is illustrated, but be not limited to the defects detection Flame Image Process, also can (for example carry out various Flame Image Process, the deformation process of image, noise removing are handled, image transformation is handled, compression adjustment is handled etc.) device in, the relevant parameter value of Flame Image Process that adopts aforesaid way to be adjusted to carry out in the device.As an example, for describing in the situation of checking employing aforesaid way in the uneven device (asymmetric checking device) that on the liquid crystal pattern, generates.Wherein, " inhomogeneous " of so-called liquid crystal pattern is meant, for example in the color filter of the display part that becomes liquid crystal, in the even part of the density unevenness that should produce in the shades of colour for the R of uniform concentration (redness), G (green), B (blueness) in whole.
In asymmetric checking device, handle the non-accidental inhomogeneous part that produces on the inspection object of appointment color filter etc. by the inhomogeneous detection that puts rules into practice.Wherein, device detects for example to handle and carries out in the following manner.At first, the view data of checking object is carried out data parsing, be extracted in check generate in the object inhomogeneous (for example, this processing has by carrying out combination that concentration analysis, filtration treatment, compression are strengthened handling, the high speed Fourier transform (FastFourier Transform, FFT) and the inhomogeneous extraction algorithm of 2 value processing etc. carry out).Then, the inhomogeneous part that evaluation extracts, detect non-accidental inhomogeneous part (for example, this processing can have the inhomogeneous evaluation algorithm of the judgment processing relatively judged with the theoretical value that adds bulk processing, refinement processing, aspect ratio, area, position etc. etc. to carry out by carrying out combination).Here, uneven extraction precision or inhomogeneous detection sensitivity are stipulated by the parameter that each processing relates to.Therefore, inhomogeneous in order to extract with suitable precision, and inhomogeneous with suitable (that is, wishing) sensitivity detection according to the user, need suitably to adjust and handle the parameter value that relates to.
The asymmetric checking device that this variation relates to has the function portion that the handling part conduct of carrying out inhomogeneous detection processing is equivalent to defects detection handling part 130 (Fig. 2).This handling part, by the view data to the inspection object, the inhomogeneous detection that puts rules into practice is handled, and detects the inhomogeneous part that produces in this view data, generate the Flame Image Process that shows this detected inhomogeneous partial data (that is result data D2) with feature mode.
For other function portion relevant with parameter adjustment, identical with above-mentioned embodiment (with reference to Fig. 2), target data generating unit 140 receives revision directive to this result data D2 from the operator, generate simultaneously and make this command content be reflected in view data (target data D3) among the result data D2, and parameter adjustment handling part 170 determines to adjust the adjusted value of image parameter.Thus, can carry out inhomogeneous detection and handle the parameter that relates to.
＜the 3 variation 〉
In the respective embodiments described above, be illustrated, also can carry out the parameter adjustment that the various processing to the data except view data (for example vector data) relate to for the situation of the adjustment of carrying out the parameter that Flame Image Process relates to.The example of the situation of the parameter adjustment that relates to as the processing of carrying out vector data describes for following situation, that is, from containing the determination data of noise, utilize the tentative calculation mistake to obtain in the operation of the data of specifying determinand exactly and adopt aforesaid way.
For example, making determination data is the determining film thickness data (data of obtaining by the thickness that utilizes the beam splitting type film thickness gauge to measure to form in its surface the substrate of groove (pattern)) of substrate.Usually, in determination data, can contain noise.These determining film thickness data also contain noise effect, are difficult to directly read from determination data the width or the degree of depth of actual groove.Therefore, use the method for following simulation to infer the width of groove or the operation of the degree of depth.
At first, generate the vector data (that is, the vector data of the shape of performance groove) of the suitable width of regulation and the degree of depth, this vector data is carried out the beam split simulation.That is to say, be created on imaginaryly and utilize the beam splitting type film thickness gauge, measure by resulting measured value under the situation of the groove of this vector data performance (below, be referred to as " imaginary data ").If should the imagination data and the state consistency of having removed noise, then have only the vector that can access these imagination data, be only the width that correctly shows groove or the data of the degree of depth.Therefore, the vector data that the operator seeks to draw these imagination data according to the tentative calculation mistake (more particularly, compare by the imaginary data and the measurement data that will obtain,, the vector data shape is changed to eliminate the mode of the different piece that between two data, produces.Then,, regenerate imaginary data, imaginary data and the measurement data that obtains compared, repeat above-mentioned operation and search) according to vector data after changing.
In this operation, can adopt above-mentioned parameter to adjust function.In this variation, be provided with the handling part (vector data generation handling part) of the generation processing of carrying out vector data, as the function portion that is equivalent to defects detection handling part 130 (Fig. 2).This handling part generates the vector data that performance has the groove of the width of regulation and the degree of depth.Wherein, the width of the groove of performance or the degree of depth generate the parameter value regulation that handling part provides by this vector data, in this variation, make this parameter for adjusting object data.By vector data is generated the vector data that handling part generates, the beam split of stipulating simulation can obtain imaginary data.In this variation, the imaginary data that herein obtain are result data D2.
For other relevant function portion of parameter adjustment, identical with above-mentioned embodiment (with reference to Fig. 2).That is to say that target data generating unit 140 receives revision directive to result data D2 from the operator, generate simultaneously this command content is reflected to data (target data D3) among the result data D2.In this variation, the operator carries out the correction to result data D2, so that target data D3 that generates and the state consistency of eliminating noise from determination data.
And, parameter adjustment handling part 170, obtain can access the result data D2 consistent with target data D3 parameter value as adjusted value.That is to say, utilize parameter adjustment handling part 170, obtain the parameter value that can generate suitable vector data (that is, can with the vector data of the imaginary data of the state consistency of from determination data, eliminating noise).The vector data that uses this parameter value to generate becomes the width of correct performance groove or the data of the degree of depth.Thus, by the adjustment function of operation parameter, can be easily and carry out exactly inferring the width of actual groove or the operation of the degree of depth according to determination data.
＜the 4 variation 〉
The respective embodiments described above are, as also illustrating in the 3rd variation, are the effective technologies in the field of using simulation.For example, in the manufacturing process of various products, generate shape after the processing by the shape of material of simulation before according to processing sometimes, whether the shape after confirming to process is the operation of target shape.
In this operation, can adopt above-mentioned parameter to adjust function.In this variation, be provided with handling part (material shape generation handling part) as the function portion that is equivalent to defects detection handling part 130 (Fig. 2), its generation of carrying out the shape data of the material shape before the expression processing is handled.This handling part generates the shape data of the material shape of performance regulation.Wherein, the shape of performance generates the parameter value regulation that handling part provides by this material shape, in this variation, with this parameter as adjusting image parameter.Especially, the density state of the grid of performance shape also can be used as and adjusts one of image parameter.By material shape being generated the shape data that handling part generates, the simulation of the processing operation of stipulating can obtain to represent to process the shape data of the shape after this material shape.In this variation, the data that herein obtain will become result data D2.
For other function portion relevant, identical with above-mentioned embodiment (with reference to Fig. 2) with parameter adjustment.That is, target data generating unit 140 receives revision directive to result data D2 from the operator, generates simultaneously this command content is presented at data (target data D3) among the result data D2.In this variation, the operator carries out the correction to result data D2 in the mode that the target data D3 itself that generates represents the shape after the desired processing.
And parameter adjustment handling part 170 obtains the parameter value that can access the result data D2 consistent with target data D3, as adjusted value.That is to say, acquisition can be generated the parameter value of the data of suitable material shape (that is, can obtain the material shape of the desired machining shape of operator).Thus, by the adjustment function of operation parameter, can be easily and infer the material shape that is used to obtain desired machining shape exactly.
＜the 5 variation 〉
As the function portion of the adjustment parameter that in above-mentioned each embodiment, illustrates, the structure (parameter evaluation portion (omitting diagram)) of estimating the adjusted value that obtains can also be set.
Whether the adjusted value that parameter evaluation portion evaluating is adjusted the definite adjustment image parameter of handling part 170 (Fig. 2) is suitable value.For example, compare the data (net result data) and target data D3 under the adjusted value of determining, carrying out the defects detection algorithm and obtain, calculate the area occupation ratio of the uniform portion of two data, obtain the evaluation of estimate of this value as adjusted value.That is to say, we can say that evaluation of estimate is high more, then the degree that the data that obtain under the adjusted value that obtains are consistent with target data D3 is high more.The evaluation of estimate that obtains for example, can show on display part 14, informs the operator.In addition, be lower than in evaluation of estimate under the situation of setting, the structure of final fine setting parameter value can also be set, also can make parameter adjustment handling part 170 re-execute the computing of adjusted value.
＜other variation 〉
In the respective embodiments described above, in checking object data D1, represent defect area T1, T2, adopt mode thus, but be not limited thereto in the mode in the mode display defect zone of feature with the mode display defect zone of feature with peony.For example, also can adopt than appropriate area deeper mode, the flicker display defect zone in display defect zone mode, show the mode of appropriate area etc. with the color (for example, light gray) of regulation.
In addition, in the respective embodiments described above, only under the on all four situation of defect area part of the defect area of result data D2 part and target data D3, the tentative calculation value that judgement is adopted when generating this result data D2, for adjusting the appropriate value of image parameter, even but the part of the defect area of result data D2 is not quite identical with the part of the defect area of target data D3, under the situation of inconsistent degree in the permissible range of regulation, also can think two kinds of data consistents, the tentative calculation value that judgement is adopted when generating this result data D2 is for adjusting the appropriate value of image parameter.For example, can be in two data under the situation of area less than setting of inconsistent defect area, think two data unanimities, judge that this tentative calculation value is for adjusting the appropriate value of image parameter.
In addition, in the respective embodiments described above, become in definite processing of adjusted value, parameter adjustment handling part 170 is when making parameter value variation, make setting range begin to increase gradually (the step S31 of Figure 10), but the mode of parameter value variation is not limited thereto from minimum value by the value of adjusting range regulation parameter.For example, can reduce in order, also can be on one side begin according to the increase and decrease of unit change amount from the median of setting range from the maximal value of setting range, change on one side.In addition, can also make the value of adjusting preceding setting make its variation as initial value.
In addition, in the 1st embodiment, select by parameter value whether suitably judging part 171 be judged as appropriate value the median of tentative calculation value as adjusted value (the step S41 of Figure 10), but determine that by appropriate value the median mode is not limited thereto.For example, also tentative calculation value (adjusting image parameter is under a plurality of situation, is the combination of the parameter value that the is judged as appropriate value) tabulation on display part 14 that is judged as appropriate value can be shown, make the user select some appropriate values.
In addition, in the 2nd embodiment, obtain 3 information (area S, number A, Luminance Distribution H), but also can adopt at least 1 structure obtaining these characteristic quantities as characteristic quantity as characteristic quantity.In addition, the information of obtaining as characteristic quantity is not limited thereto.For example, also can obtain the user specifies and pulls the zone as characteristic quantity when revising subject area U.In addition, adjust image parameter determination portion 253, also can specify when becoming the parameter of adjusting object, the judged result that obtains according to the characteristic specified amount is weighted, calculate the shared ratio of each parameter in comprehensive judged result of storing.
In addition, in the 2nd embodiment, the adjustment image parameter specifying part 150 that the flaw detection apparatus 1 that adopts replacement the 1st embodiment to relate to has, the structure of adjusting image parameter judging part 250 is set, the structure of adjusting image parameter specifying part 150 and adjusting image parameter judging part 250 is set simultaneously but also can adopt.Under this situation, also the structure that can adopt the user to select promptly, is to utilize adjustment image parameter judging part 250 to adjust image parameter automatically to specify, and still the selection input that receives the operator by adjustment image parameter specifying part 150 is carried out.
In addition, in the 2nd embodiment, adjust image parameter determination portion 253 from the parameter of judged result, occupying higher proportion, successively 2 parameters are defined as adjusting image parameter, but also can be according to the height order of proportion, select 1 or, be defined as adjusting image parameter more than or equal to 2 parameter.
Claims (26)
1. image processing apparatus is characterized in that having:
Graphics processing unit, the Flame Image Process that it is stipulated;
The target image generation unit, it receives the revision directive to the result images that generates as aforementioned processing result image, generates aforementioned revision directive to be reflected in the aforementioned result image and the target image that obtains; And
The parameter adjustment unit, it adjusts the value of the parameter relevant with aforementioned Flame Image Process, so that the result images that generates as aforementioned processing result image is consistent with aforementioned target image.
2. image processing apparatus as claimed in claim 1 is characterized in that,
Aforementioned graphics processing unit carries out following Flame Image Process, promptly, the inspection object images that benchmark image is obtained with this benchmark image of output compares, the defect part that detection produces in aforementioned inspection object images, generation shows the result images of this detected defect part with the display format different with non-defect part
The parameter relevant with aforementioned Flame Image Process, the detection sensitivity of regulation aforementioned disadvantages part.
3. image processing apparatus as claimed in claim 1 or 2 is characterized in that,
Aforementioned target image generation unit is presented at the aforementioned result image in the display frame, from the input of aforementioned display frame reception to the revision directive of this result images.
4. image processing apparatus as claimed in claim 3 is characterized in that,
The aforementioned parameters adjustment unit makes the value of aforementioned parameters change on one side, judge that on one side whether the result images that obtains is consistent with aforementioned target image, is defined as the adjusted value of aforementioned parameters can obtain the parameter value of the result images consistent with aforementioned target image under each parameter value.
5. image processing apparatus as claimed in claim 4 is characterized in that,
The aforementioned parameters adjustment unit is defined as the median of following set with the adjusted value of aforementioned parameters, and this set is made of the parameter value that can obtain the result images consistent with aforementioned target image.
6. image processing apparatus as claimed in claim 4 is characterized in that,
The aforementioned parameters adjustment unit is defined as following parameter value with the aforementioned parameters adjusted value, and this parameter value is the parameter value in the time of can obtaining the result images consistent with aforementioned target image at first in the process that the aforementioned parameters value is changed.
7. image processing apparatus as claimed in claim 4 is characterized in that,
Also have the setting range designating unit, it specifies the scope that the aforementioned parameters value is changed according to the setting value of input.
8. image processing apparatus as claimed in claim 4 is characterized in that,
Also have the variable quantity designating unit, it specifies the variable quantity when the aforementioned parameters value is changed according to the setting value of input.
9. image processing apparatus as claimed in claim 3 is characterized in that,
The aforementioned parameters adjustment unit is from the combination of the parameter value more than 1 or 1, utilize genetic algorithm, extraction can obtain the parameter value combination of the result images consistent with aforementioned target image, the adjusted value of aforementioned parameters is defined as constituting each value of this combination that extracts.
10. image processing apparatus as claimed in claim 9 is characterized in that,
Also have the setting range designating unit, it specifies the scope that the aforementioned parameters value is changed according to the setting value of input.
11. image processing apparatus as claimed in claim 9 is characterized in that,
Also have the variable quantity designating unit, it specifies the variable quantity when the aforementioned parameters value is changed according to the setting value of input.
12. image processing apparatus as claimed in claim 1 or 2 is characterized in that,
Have the image parameter of adjustment designating unit, it, is specified as the parameter of adjusting object from the parameter that aforementioned Flame Image Process relates to according to the characteristic quantity of the regulation that obtains from aforementioned revision directive.
13. a parameter regulation means is characterized in that having:
Target image generates operation, and it receives revision directive to the result images that generates as carrying out the result of Flame Image Process from the operator, generates aforementioned revision directive to be reflected in the aforementioned result image and the target image that obtains; And
The parameter adjustment operation, it adjusts the value of the parameter relevant with aforementioned Flame Image Process, so that the result images that generates as carrying out the result of aforementioned Flame Image Process is consistent with aforementioned target image.
14. a data processing equipment is characterized in that having:
Data processing unit, it carries out predetermined data and handles;
The target data generation unit, it receives the revision directive of the result data that the result who handles as aforementioned data is obtained, and generation makes aforementioned revision directive be reflected in the target data that obtains in the aforementioned result data; And
The parameter adjustment unit, it adjusts the value of handling relevant parameter with aforementioned data, so that the data of obtaining as the aforementioned data result are consistent with aforementioned target data.
15. data processing equipment as claimed in claim 14 is characterized in that,
The aforementioned data processing unit carries out following Flame Image Process, promptly, the inspection object images that benchmark image is obtained with this benchmark image of output compares, the defect part that detection produces in aforementioned inspection object images, generation shows the view data of this detected defect part with the display format different with non-defect part, obtain as the aforementioned result data
Handle relevant parameter with aforementioned data, the detection sensitivity of regulation aforementioned disadvantages part.
16. as claim 14 or the described data processing equipment of claim 15, it is characterized in that,
Aforementioned target data generation unit in display frame, receives input to the revision directive of this result data from aforementioned display frame with the aforementioned result data presentation.
17. data processing equipment as claimed in claim 16 is characterized in that,
The aforementioned parameters adjustment unit makes the value of aforementioned parameters change on one side, judge on one side whether the result data that obtains is consistent with aforementioned target data under each parameter value, with the adjusted value of aforementioned parameters, be defined as to obtain the parameter value of the result data consistent with aforementioned target data.
18. data processing equipment as claimed in claim 17 is characterized in that,
The aforementioned parameters adjustment unit is defined as the median of following set with the adjusted value of aforementioned parameters, and this set is made of the parameter value that can obtain the result data consistent with aforementioned target data.
19. data processing equipment as claimed in claim 17 is characterized in that,
The aforementioned parameters adjustment unit is defined as following parameter value with the adjusted value of aforementioned parameters, the parameter value of this parameter value when obtaining the result data consistent with aforementioned target data at first in the process that the aforementioned parameters value is changed.
20. data processing equipment as claimed in claim 17 is characterized in that,
Also have the setting range designating unit, it specifies the scope that the aforementioned parameters value is changed according to the setting value of input.
21. data processing equipment as claimed in claim 17 is characterized in that,
Also have the variable quantity designating unit, it specifies the variable quantity when the aforementioned parameters value is changed according to the setting value of input.
22. data processing equipment as claimed in claim 16 is characterized in that,
The aforementioned parameters adjustment unit, from the combination of the parameter value more than 1 or 1, utilize genetic algorithm, extraction can draw the parameter value combination of the result data consistent with aforementioned target data, the adjusted value of aforementioned parameters is defined as constituting each value of this combination that extracts.
23. data processing equipment as claimed in claim 22 is characterized in that,
Also have the setting range designating unit, it specifies the scope that the aforementioned parameters value is changed according to the setting value of input.
24. data processing equipment as claimed in claim 22 is characterized in that,
Also have the variable quantity designating unit, it specifies the variable quantity when the aforementioned parameters value is changed according to the setting value of input.
25. as claim 14 or 15 described data processing equipments, it is characterized in that,
Have the image parameter of adjustment designating unit, it, is specified as the parameter of adjusting object from handling the relevant parameter with aforementioned data according to the characteristic quantity of the regulation of obtaining from aforementioned revision directive.
26. a parameter regulation means is characterized in that having:
Target data generates operation, and it receives revision directive to the result data that obtains as carrying out The results of data processing from the operator, generates to make aforementioned revision directive be reflected in the target data that obtains in the aforementioned result data; And
The parameter adjustment operation, it adjusts the value of handling relevant parameter with aforementioned data, so that the data of obtaining as the result who carries out the aforementioned data processing are consistent with aforementioned target data.
Applications Claiming Priority (3)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| JP2006239820 | 2006-09-05 | ||
| JP2006239820 | 2006-09-05 | ||
| JP2007206694 | 2007-08-08 |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN101140619A true CN101140619A (en) | 2008-03-12 |
Family
ID=39192566
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CNA2007101458854A Pending CN101140619A (en) | 2006-09-05 | 2007-09-04 | Image processing device, data processing device and parameter adjusting method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN101140619A (en) |
Cited By (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103606160A (en) * | 2013-12-02 | 2014-02-26 | 苏州比特速浪电子科技有限公司 | Image processing device and image processing method |
| CN104063836A (en) * | 2013-03-15 | 2014-09-24 | 株式会社理光 | Image Inspection Device, Image Inspection System, And Image Inspection Method |
| CN106525861A (en) * | 2016-10-13 | 2017-03-22 | 广州视源电子科技股份有限公司 | AOI test parameter adjusting method and device |
| CN107024541A (en) * | 2015-10-08 | 2017-08-08 | 株式会社日立电力解决方案 | Defect detecting method and its device |
| CN109462999A (en) * | 2016-05-23 | 2019-03-12 | Ani有限公司 | Vision inspection method by data balancing based on study and the vision inspection device using it |
| CN109873922A (en) * | 2019-02-20 | 2019-06-11 | 广州视源电子科技股份有限公司 | Method and device for adjusting overscan parameters, display equipment and storage medium |
| CN110658199A (en) * | 2018-06-29 | 2020-01-07 | 欧姆龙株式会社 | Inspection result presentation device, inspection result presentation method, and storage medium |
| CN114113099A (en) * | 2020-08-28 | 2022-03-01 | 致茂电子(苏州)有限公司 | Method for automatically adjusting electronic equipment |
-
2007
- 2007-09-04 CN CNA2007101458854A patent/CN101140619A/en active Pending
Cited By (13)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104063836A (en) * | 2013-03-15 | 2014-09-24 | 株式会社理光 | Image Inspection Device, Image Inspection System, And Image Inspection Method |
| CN104063836B (en) * | 2013-03-15 | 2018-01-30 | 株式会社理光 | Image inspection equipment, image review systems and image checking method |
| CN103606160A (en) * | 2013-12-02 | 2014-02-26 | 苏州比特速浪电子科技有限公司 | Image processing device and image processing method |
| CN107024541A (en) * | 2015-10-08 | 2017-08-08 | 株式会社日立电力解决方案 | Defect detecting method and its device |
| CN109462999B (en) * | 2016-05-23 | 2021-05-07 | Ani有限公司 | Visual inspection method based on learning through data balance and visual inspection device using same |
| CN109462999A (en) * | 2016-05-23 | 2019-03-12 | Ani有限公司 | Vision inspection method by data balancing based on study and the vision inspection device using it |
| CN106525861A (en) * | 2016-10-13 | 2017-03-22 | 广州视源电子科技股份有限公司 | AOI test parameter adjusting method and device |
| CN106525861B (en) * | 2016-10-13 | 2019-09-17 | 广州视源电子科技股份有限公司 | AOI test parameter adjusting method and device |
| CN110658199A (en) * | 2018-06-29 | 2020-01-07 | 欧姆龙株式会社 | Inspection result presentation device, inspection result presentation method, and storage medium |
| CN109873922A (en) * | 2019-02-20 | 2019-06-11 | 广州视源电子科技股份有限公司 | Method and device for adjusting overscan parameters, display equipment and storage medium |
| CN109873922B (en) * | 2019-02-20 | 2021-01-19 | 广州视源电子科技股份有限公司 | Method and device for adjusting overscan parameters, display equipment and storage medium |
| CN114113099A (en) * | 2020-08-28 | 2022-03-01 | 致茂电子(苏州)有限公司 | Method for automatically adjusting electronic equipment |
| CN114113099B (en) * | 2020-08-28 | 2024-10-11 | 致茂电子(苏州)有限公司 | Method for automatically adjusting electronic equipment |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| JP2008089574A (en) | Image processor, data processor, parameter adjustment method, and program | |
| CN101140619A (en) | Image processing device, data processing device and parameter adjusting method | |
| RU2680765C1 (en) | Automated determination and cutting of non-singular contour of a picture on an image | |
| CN104075666B (en) | Enhanced edge detection tool for irregular surface edge | |
| CN105849886B (en) | Method and apparatus for analyzing the shape of wafer | |
| CN1940912B (en) | Document creation system, document creation method | |
| EP3762795B1 (en) | Method, device, system and program for setting lighting condition and storage medium | |
| EP2179396B1 (en) | Toner-based noise reduction in electrostatography | |
| CN103196914A (en) | Visual inspection device and visual inspection method | |
| US20200364913A1 (en) | User guided segmentation network | |
| US20070127783A1 (en) | Image processing apparatus, method and program for controlling flesh color of image | |
| US20230206416A1 (en) | Computer-implemented method for quality control of a digital image of a sample | |
| US20080232652A1 (en) | Device, method and computer readable recording medium containing program for evaluating image | |
| CN117635523A (en) | Microscope system and method for generating machine learning model for processing microscope data | |
| CN107094219B (en) | Image processing apparatus and image processing method | |
| US6675714B2 (en) | Ink and water supply controller in printing machine, printing system with such controller, and program therefor | |
| US20050280845A1 (en) | Color correction method with improved image translation accuracy | |
| Fukui et al. | Analysis of chromosome information | |
| JPH1063859A (en) | Method for evaluating picture quality and device therefor | |
| JPH10164281A (en) | Picture evaluating method and device | |
| EP2187615A2 (en) | FM-Halftoning with surrounding-influenced quantization and lazy propagation | |
| JP4793648B2 (en) | Image correction unit and image correction program | |
| JP4941593B2 (en) | Material surface roughness detection program, computer-readable recording medium having the program recorded thereon, and material surface roughness detection device | |
| JP6815431B2 (en) | Painted surface inspection equipment and programs | |
| EP2437481B1 (en) | Preferred hue selection method for optimizing color |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C02 | Deemed withdrawal of patent application after publication (patent law 2001) | ||
| WD01 | Invention patent application deemed withdrawn after publication |
Open date: 20080312 |