CN101569176B - Signal processing device - Google Patents
Signal processing device Download PDFInfo
- Publication number
- CN101569176B CN101569176B CN200780048123.0A CN200780048123A CN101569176B CN 101569176 B CN101569176 B CN 101569176B CN 200780048123 A CN200780048123 A CN 200780048123A CN 101569176 B CN101569176 B CN 101569176B
- Authority
- CN
- China
- Prior art keywords
- data
- signal
- processing
- signal data
- pixel
- 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.)
- Expired - Fee Related
Links
- 238000012545 processing Methods 0.000 title claims abstract description 227
- 238000000034 method Methods 0.000 claims abstract description 66
- 230000008569 process Effects 0.000 claims abstract description 35
- 230000008859 change Effects 0.000 claims abstract description 34
- 238000012804 iterative process Methods 0.000 claims abstract 7
- 230000003321 amplification Effects 0.000 claims description 24
- 230000006866 deterioration Effects 0.000 claims description 24
- 238000003199 nucleic acid amplification method Methods 0.000 claims description 24
- 230000003252 repetitive effect Effects 0.000 claims 3
- 230000000593 degrading effect Effects 0.000 claims 1
- 238000011946 reduction process Methods 0.000 claims 1
- 238000003672 processing method Methods 0.000 abstract description 8
- 238000012958 reprocessing Methods 0.000 description 96
- 239000011435 rock Substances 0.000 description 18
- 238000011084 recovery Methods 0.000 description 15
- 230000015572 biosynthetic process Effects 0.000 description 10
- 238000012360 testing method Methods 0.000 description 9
- 238000001514 detection method Methods 0.000 description 8
- 238000012546 transfer Methods 0.000 description 8
- 238000009826 distribution Methods 0.000 description 7
- 230000003287 optical effect Effects 0.000 description 6
- 230000009466 transformation Effects 0.000 description 5
- 238000005070 sampling Methods 0.000 description 4
- 230000004044 response Effects 0.000 description 3
- 238000012163 sequencing technique Methods 0.000 description 3
- 244000287680 Garcinia dulcis Species 0.000 description 2
- 230000001133 acceleration Effects 0.000 description 2
- 230000004075 alteration Effects 0.000 description 2
- 230000008901 benefit Effects 0.000 description 2
- 238000013527 convolutional neural network Methods 0.000 description 2
- 239000000284 extract Substances 0.000 description 2
- 238000005457 optimization Methods 0.000 description 2
- 239000004065 semiconductor Substances 0.000 description 2
- 230000007704 transition Effects 0.000 description 2
- TWDJIKFUVRYBJF-UHFFFAOYSA-N Cyanthoate Chemical compound CCOP(=O)(OCC)SCC(=O)NC(C)(C)C#N TWDJIKFUVRYBJF-UHFFFAOYSA-N 0.000 description 1
- 241001269238 Data Species 0.000 description 1
- 239000000654 additive Substances 0.000 description 1
- 230000000996 additive effect Effects 0.000 description 1
- 230000015556 catabolic process Effects 0.000 description 1
- 230000000295 complement effect Effects 0.000 description 1
- 239000012141 concentrate Substances 0.000 description 1
- 238000000354 decomposition reaction Methods 0.000 description 1
- 238000006731 degradation reaction Methods 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 239000006185 dispersion Substances 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 238000002474 experimental method Methods 0.000 description 1
- 230000002349 favourable effect Effects 0.000 description 1
- 230000008676 import Effects 0.000 description 1
- 238000003780 insertion Methods 0.000 description 1
- 230000037431 insertion Effects 0.000 description 1
- 230000007246 mechanism Effects 0.000 description 1
- 229910044991 metal oxide Inorganic materials 0.000 description 1
- 150000004706 metal oxides Chemical class 0.000 description 1
- 239000000203 mixture Substances 0.000 description 1
- 238000012544 monitoring process Methods 0.000 description 1
- 238000005096 rolling process Methods 0.000 description 1
- 230000035945 sensitivity Effects 0.000 description 1
- 238000000926 separation method Methods 0.000 description 1
- 238000004904 shortening Methods 0.000 description 1
- 230000005236 sound signal Effects 0.000 description 1
- 238000003860 storage Methods 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
- 238000011282 treatment Methods 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T5/00—Image enhancement or restoration
- G06T5/20—Image enhancement or restoration using local operators
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N23/00—Cameras or camera modules comprising electronic image sensors; Control thereof
- H04N23/80—Camera processing pipelines; Components thereof
Landscapes
- Engineering & Computer Science (AREA)
- Multimedia (AREA)
- Signal Processing (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Image Processing (AREA)
- Editing Of Facsimile Originals (AREA)
- Facsimile Image Signal Circuits (AREA)
- Studio Devices (AREA)
- Studio Circuits (AREA)
Abstract
本发明提供的信号处理装置，在复原处理时防止装置的大型化，同时，设有具有现实性的电路处理方式；信号处理装置(1)的处理部(4)，使用将原信号数据的容量进行缩小处理(步骤S203)后的缩小原信号数据，并利用变化要因信息的数据由任意的图像数据生成比较用数据；然后，进行通过利用变化要因信息的数据将成为处理对象的图像数据与比较用数据的差分的数据分配于任意的图像数据，而生成复原数据，取代任意的图像数据而使用该复原数据并重复同样的处理的重复处理(步骤S204)；然后，使用将最后的(缩小)复原数据进行放大处理(步骤S205)后的放大信号数据，进行上述的重复处理(步骤S206)；另外，也可以仅以第一重复处理(步骤S204)而结束。
The signal processing device provided by the present invention prevents the enlargement of the device during restoration processing, and at the same time, is provided with a realistic circuit processing method; the processing part (4) of the signal processing device (1) uses the capacity of the original signal data Perform the reduced original signal data after the reduction processing (step S203), and use the data of the change factor information to generate comparison data from any image data; then, compare the image data to be processed by using the change factor information The data of the data difference is allocated to arbitrary image data to generate restored data, and the restored data is used instead of arbitrary image data to repeat the same process (step S204); then, use the last (reduced) The amplified signal data after the restoration data has been amplified (step S205) is subjected to the above-mentioned iterative process (step S206); in addition, it may end with only the first iterative process (step S204).
Description
Technical field
The present invention relates to signal processing apparatus.
Background technology
As everyone knows under prior art, when signal (image) processing unit that utilizes camera etc. was taken object, deterioration took place through regular meeting in its image.As the main cause of image deterioration, rocking when exist taking, the various aberrations of optical system, the deflection of camera lens etc.
Correct owing to rock and the method for the photographs of deterioration the method that the known Method and circuits that moving lens is arranged is handled when taking.For example, as the method for moving lens, known has through detecting rocking of camera, and the camera lens of regulation is moved corresponding to this detected rocking, thus the method for correcting (with reference to patent documentation 1).In addition; Method as processing of circuit; Known have a change that utilizes angular acceleration transducer to detect the camera optical axis; The transfer function of the fringe when obtaining expression and take, and the inverse transformation of the transfer function that photographs is obtained, thereby the method (with reference to patent documentation 2) that the image of deterioration is corrected by detected angular speed etc.
In addition, except that general photographs, various images such as X-ray photograph, MIcrosope image are owing to rock or other former thereby situation that deterioration or variation take place also are well-known.
Patent documentation 1: day disclosure communique, spy open flat 6-317824 number (with reference to specification digest)
Patent documentation 2: day disclosure communique, spy open flat 11-24122 number (with reference to specification digest)
Summary of the invention
Adopt the camera that rocks rectification of patent documentation 1 record, owing to need the configuration space of the hardware of driving camera lens such as motor to cause maximization.In addition, owing to need this hardware self or drive the drive circuit of this hardware, so cost increases.In addition, under the situation of rocking rectification of patent documentation 2 records,, has following problem though can address the above problem a little.That is, setting up in theory though image restoration is carried out in the inverse transformation of the transfer function of passing through to be obtained, as practical problem, is difficult because following two reasons cause image restoration.
The first, the transfer function that is obtained for noise or rock information error etc. very a little less than, these have change slightly, very big change just takes place in transfer function values.Therefore, the correcting image through inverse transformation obtains differs greatly with the image of under the state that does not shake, taking, and in fact can not utilize.Second; Under the situation of the inverse transformation of having considered noise etc., also can adopt singular value decomposition of separating through simultaneous equations etc. to calculate the method for separating, still; The calculated value that is used for this reckoning has astronomical figure so big, and that in fact can't answer is dangerous high.
The problems referred to above that produce in the image also come across in the general various signal datas, carry out the recovery of signal through the inverse transformation of transfer function, difficulty that yes when the transfer function that obtains is incorrect, even when correct also inconvenient situation.And it is impossible under the situation that with the nature is object, obtaining very correct transfer function.
As stated, problem of the present invention is to provide a kind of maximization of when release signal, preventing locking apparatus, simultaneously, is provided with the signal processing apparatus of the processing of circuit mode with actuality.
In order to solve above-mentioned problem; Signal processing apparatus of the present invention; Data with the variation essential factor information when utilizing variation are by the original signal data that deterioration changes have taken place; The handling part of formation base signal data or similarity signal data, these basis signal data are meant the data of the signal before changing or the data of its approximate signal, and these similarity signal data are the data of the basis signal data being dwindled or amplifying back formation; Handling part is made up of the module of carrying out first reprocessing, the module of carrying out the module of processing and amplifying and carrying out second reprocessing; Wherein, The same processing that first reprocessing repeats is: according to the data that change essential factor information; To dwindle original signal data and first differential data of relatively using two data of dwindling data; Be allocated in arbitrarily signal data and generate and dwindle restored data, replace arbitrarily signal data and use this to dwindle restored data; The said original signal data of dwindling, be carried out dwindling the original signal data capacity dwindle the data after the processing; Said relatively with dwindling data, be with the data that utilize to change essential factor information by signal data arbitrarily generate to dwindle the original signal data capacity identical and make the data after the signal data deterioration arbitrarily; Processing and amplifying is that the capacity that dwindles restored data that obtains through first reprocessing of carrying out in the above-mentioned module of carrying out first reprocessing is amplified; The same processing that second reprocessing repeats is: generated by the amplifying signal data of in above-mentioned module of carrying out processing and amplifying, having carried out processing and amplifying and relatively use amplification data after making this amplifying signal data degradation; According to the data that change essential factor information; With original signal data or original signal similarity signal data and second differential data of relatively using amplification data; Be allocated in the amplifying signal data and generate the amplification restored data, replace the amplifying signal data and use this amplification restored data; Wherein, said original signal similarity signal data are that capacity with the original signal data dwindles or is enlarged into the data behind the capacity identical with relatively using amplification data.
Adopt this invention, can carry out by of the recovery of original signal data to basis signal data or similarity signal data.At this moment, almost do not have the increase on the hardware, thereby device can not maximize.In addition; In first reprocessing; Repetition is made relatively with dwindling data by signal data arbitrarily, and utilizes the data that change essential factor information, and this relatively is allocated in the processing of signal data arbitrarily with first differential data that dwindles the original signal data that dwindles data and process object; In second reprocessing; Repetition is made and is relatively used amplification data by dwindling restored data; And utilize to change the data of essential factor information, this relatively is allocated in the processing of amplifying signal data with second differential data of the original signal data of amplification data and process object or original signal similarity signal data.Because through this first reprocessing and second reprocessing; Obtain gradually and basis signal data or the approaching amplification release signal data of similarity signal data; Be provided with the signal processing apparatus of processing of circuit mode when therefore, can be formed in the recovery of signal with actuality.In addition, in first reprocessing,, but dwindle the original signal data after its capacity dwindled, therefore can improve processing speed because the object of reprocessing is not the original signal data.Therefore, can make the first reprocessing short timeization, simultaneously, if the words of the processing in the stipulated time just can repeat more frequently, thereby the recovering quality that dwindles restored data becomes good.And, carrying out second reprocessing through the restored data of dwindling of utilizing this good quality, it is good that the recovering quality of amplifying signal also becomes.
The signal processing apparatus that other inventions relate to is on the basis of foregoing invention, dwindles to handle to be, as the capacity of x, original signal data during as y, the value that makes x/y is below 0.9 at the capacity that will dwindle the original signal data.Through make this than value be below 0.9, can promptly carry out first reprocessing, therefore; Can make the first reprocessing short timeization; Simultaneously, if the words of the processing in the stipulated time just can be carried out more times, thereby the recovering quality that dwindles restored data becomes good.
The signal processing apparatus that other inventions relate to is on the basis of foregoing invention; Handling part; Only in the variation that obtains by the data that change essential factor information under the situation more than the setting, dwindle processing, first reprocessing, processing and amplifying and second reprocessing.Through the employing of this formation, only,, carry out by of the recovery of original signal data to basis signal data or similarity signal data to follow the state that dwindles processing and processing and amplifying changing under the bigger situation.Therefore, owing to have only existence to need the data of the tendency of spended time to become process object, therefore becoming more favourable aspect the time of shortening.
The signal processing apparatus that other inventions relate to is on the basis of foregoing invention; Handling part; To dwindle the original signal data and relatively with dwindling after data contrast; First differential data become setting when following or the number of repetition of first reprocessing be setting when above, finish first reprocessing; With original signal data or original signal similarity signal data with after relatively contrasting with amplification data, second differential data become setting when following or the number of repetition of second reprocessing be setting when above, finish second reprocessing.Through the employing of this formation, when carrying out by the original signal data, can make the condition that finishes first reprocessing and second reprocessing become suitable condition to the recovery of basis signal data or similarity signal data.At this, so-called " contrast " refers to two data compared.
The signal processing apparatus that other inventions relate to is on the basis of foregoing invention, and handling part will amplify restored data as basis signal data or similarity signal data and obtain.Through the employing of this formation, can access the basis signal data or the similarity signal data that first reprocessing and second reprocessing are finished with suitable condition.
The signal processing apparatus that other inventions relate to is on the basis of foregoing invention, and the number of times of second reprocessing is less than the number of times of first reprocessing.Adopt the signal processing apparatus of this formation; Emphasis carries out the first fast reprocessing of processing speed; And make the recovering quality that dwindles restored data good, and carry out second reprocessing through the restored data of dwindling of using this good quality, and can make the recovering quality of amplifying signal also become good.
The signal processing apparatus that other inventions relate to is on basis of above-mentioned each invention, with the data of the signal data as image.Employing through this formation; Under the situation of the image deterioration that takes place to cause by rocking of camera; Can from the original image that deterioration changes have taken place, the image before changing or the approximate image data of base image data or base image, perhaps the base image data are dwindled or amplify after the similar image data recovery.
In the present invention, the maximization of anti-locking apparatus simultaneously, is provided with the signal processing apparatus of the processing of circuit mode with actuality in the time of can being provided at release signal.
Description of drawings
Fig. 1 is the block diagram of the main composition of the signal processing apparatus that relates to of expression example of the present invention.
Fig. 2 is the stereoscopic figure of summary of expression signal processing apparatus shown in Figure 1, is the figure that is used to explain the allocation position of angular-rate sensor.
Fig. 3 is the figure that is used to explain the notion of dwindling processing method that the handling part that utilizes signal processing apparatus shown in Figure 1 carries out.
Fig. 4 is the figure that is used to explain the notion of the processing and amplifying method that the handling part that utilizes signal processing apparatus shown in Figure 1 carries out.
Fig. 5 is the process chart that is used to explain the handling procedure that the common processing section of each reprocessing that the handling part that utilizes signal processing apparatus shown in Figure 1 carries out relates to.
Fig. 6 is the figure that is used to explain the notion of processing method shown in Figure 5.
Fig. 7 is to be the figure that example is explained common processing usefulness shown in Figure 5 particularly to rock, and is the table of the concentration of energy when representing not shake.
Fig. 8 is to be the figure that example is explained common processing usefulness shown in Figure 5 particularly to rock, and is the sketch map of the view data when not shaking.
Fig. 9 is to be the figure that example is explained common processing usefulness shown in Figure 5 particularly to rock, and is the sketch map that the energy when shaking disperses.
Figure 10 is to be the figure that example is explained common processing usefulness shown in Figure 5 particularly to rock, and is to be used to explain the figure that is generated the situation of relatively using data by image arbitrarily.
Figure 11 is to be the figure that example is explained common processing usefulness shown in Figure 5 particularly to rock, and is to be used to explain will relatively compare with data and the fuzzy original image that becomes process object and generate the figure of the situation of first differential data.
Figure 12 is to be the figure that example is explained common processing usefulness shown in Figure 5 particularly to rock, and is to be used for explaining through the data of difference being distributed and making an addition to the figure that image arbitrarily generates the situation of restored data.
Figure 13 is to be the figure that example is explained common processing usefulness shown in Figure 5 particularly to rock; Be to be used to explain by the restored data that is generated generate the new data of relatively using, these data and the fuzzy original image that becomes process object are compared and generate the figure of situation of the data of difference.
Figure 14 is to be the figure that example is explained common processing usefulness shown in Figure 5 particularly to rock, and is to be used to explain the figure that the first newly-generated differential data is distributed and generates the situation of new restored data.
Figure 15 is the whole flow chart of each flow process of handling that expression utilizes the handling part of signal processing apparatus shown in Figure 1 to carry out.
Symbol description
1 signal processing apparatus
2 shoot parts (acceptance division)
3 control system portions
4 handling parts
5 recording portion
6 test sections
7 essential factor information storing section
G is by changing the variation function that essential factor information is calculated
Img original digital image data (image of shooting)
1
0+nRestored data
Img base image data (data of the image before changing)
I
0View data arbitrarily
H destination image data (at last obtain dwindle restored data or base image data)
J becomes the view data (original digital image data, dwindle original digital image data or enlarged image data) of process object
Embodiment
Below, with reference to accompanying drawing the signal processing apparatus 1 that example of the present invention relates to is described.In addition; This signal processing apparatus 1 is the civilian camera as image processing apparatus; But; Also can be used as monitoring camera, TV camera, perhaps also go in image diagnosing system that microscope, binoculars and then NMR take etc. etc., the instrument beyond the camera with other purposes such as camera, portable video camera, endoscope cameras.
The summary of the formation of Fig. 1 expression signal processing unit 1.Signal processing apparatus 1 is provided with: takes the shoot part 2 of personage's etc. image, drives the control system portion 3 of this shoot part 2, and the handling part 4 that the image that shoot part 2 is taken is handled.In addition; The signal processing apparatus 1 that this example relates to; And then be provided with: record is processed the record images portion 5 that portion 4 handles; Constitute, detect the variation essential factor detection of information portion 6 that becomes the main cause that image deterioration etc. changes by angular-rate sensor etc., and preserve the essential factor information storing section 7 that makes the variation essential factor information that image deterioration etc. takes place.In addition, at signal processing apparatus 1 as the device beyond the image processing apparatus and under the suitable situation, shoot part 2 becomes the acceptance division 2 (below, suitably shoot part 2 and acceptance division were opened use in 2 minutes) of the various input signals that receive voice signal etc.
Handling part 4 is made up of image processor, and is made up of the such hardware of ASIC (Application Specific Integr ated Circuit, application-specific integrated circuit (ASIC)).Handling part 4, the sampling frequency (sampling frequency) that the vibration detection of rocking of detecting etc. is used takes place, and simultaneously, this sampling frequency is supplied to test section 6.In addition, the beginning and the end of handling part 4 control vibration detection.And then handling part 4 dwindles the dwindling processing, will dwindle processing and amplifying, the 1st and second reprocessing that the capacity of restored data amplifies of capacity (capacity value) of original signal data.And then, handling part 4, the degree of the variation that obtains according to the data by test section 6 detected variation essential factor information judges whether to dwindle processing, first reprocessing, processing and amplifying and second reprocessing.In addition, at signal processing apparatus 1 and under the suitable situation, can change the reception sensitivity of acceptance division 2 according to the size of input signal etc. as the device beyond the image processing apparatus.
In addition, in this handling part 4, also there is the relatively situation of the original digital image data (original signal data) etc. on the basis when dwindling data and relatively using amplification data of generation that becomes of preserving.And then handling part 4 also can be used as the formation of utilizing software to handle, rather than constitutes as the such hardware of ASIC.Recording portion 5 is made up of semiconductor memory, still, also can adopt the magnetic recording system of hard drive etc. or use the optical recorder etc. of DVD etc.
Essential factor information storing section 7 is the variation essential factor recording of information portions that preserve the variation essential factor that makes Signal Degrade etc., wherein, changes essential factor information and is the aberration of for example optical system and/or test section 6 detected vibrations, or the some transform calculated in view of the above etc.Be recorded in the variation essential factor information of essential factor information storing section 7; For example from the original digital image data (view data that deterioration changes has taken place) of the original signal data that become up-to-date shooting to the base image data that become the basis signal data (view data or its approximate data before changing) or become the base image data are dwindled or amplify after the recovery processing of similar image data of similarity signal data the time, in handling part 4, be utilized.In addition; At signal processing apparatus 1 and under the suitable situation as the device beyond the image processing apparatus; Because the problem that also exists temperature, humidity etc. that the characteristic of receiving feature or the entire system of acceptance division 2 is changed; Therefore can detection record these, and utilize as changing essential factor information.In addition, also can response characteristic function of impulse response of system etc., the system known in advance etc. be stored in essential factor information storing section 7.
At this, original digital image data is restored period of processing to base image data or similar image data, the period that postpones in the time of can being power-off for shooting, when the running rate of handling part 4 is low etc., from the period of taking original image.In this case, be stored in the original digital image data of recording portion 5 and be stored in the variation essential factor information about this original digital image data (some transform etc.) of essential factor information storing section 7, preserved chronically with associated state respectively.Be the burden of handling part 4 can alleviate the shooting of following various processing time from the advantage that postpones the period of having taken original image the period of like this, the recovery of carrying out original digital image data being handled.
Then, the summary of the main processing (dwindling processing, processing and amplifying, each reprocessing) that the signal processing apparatus 1 that this example that as above, constitutes is related to reference to accompanying drawing carries out describes, and simultaneously, the handling process of integral body is described.
(dwindling processing)
Fig. 3 representes to dwindle the summary of processing.A unit of the foursquare grid among the figure, of the pixel of expression composing images data.The left side of arrow is the pixel of dwindling before handling, and the right side is the pixel of dwindling after the processing.Fig. 3 (A) expression will form the example that dwindles processing that mean value foursquare adjacent four pixels, each pixel data is replaced into the pixel data of a pixel as a whole.Minification under this situation is 25%.Fig. 3 (B) expression will form the example that dwindles processing that mean value foursquare adjacent 16 pixels, each pixel data is replaced into the pixel data of a pixel as a whole.Minification under this situation is 6.25%.
Fig. 3 (C) expression will form the example that dwindles processing that foursquare adjacent nine pixels form the pixel data of four pixels as a whole.Minification under this situation is about 44.4%.This method of dwindling processing is; The mean value of the pixel data of pixel A, B, D and E is replaced into the pixel data of pixel a; The mean value of the pixel data of pixel B, C, E and F is replaced into the pixel data of pixel b; The mean value of the pixel data of pixel D, E, G and H is replaced into the pixel data of pixel c, the mean value of the pixel data of pixel E, F, H and I is replaced into the pixel data of pixel d.
Fig. 3 just extracts the part of view data out and the summary of representing to dwindle processing.Be actually all pixels of view data are carried out the processing of dwindling shown in Figure 3.In the signal processing apparatus 1 that this example relates to, adopt the processing of dwindling shown in Fig. 3 (C).Carry out following explanation as prerequisite.In addition, dwindle to handle and be, utilize the pixel data value of the pixel that constitutes original digital image data to generate a pixel, replace these a plurality of pixels and use the processing of this pixel by a plurality of pixels.At this; The method of " utilizing the pixel data value of the pixel that constitutes original digital image data "; Be not limited to ask the method for mean value of the pixel data of a plurality of pixels, also can be method that the value of a pixel data in a plurality of pixels is intactly utilized, the method for the mean value of the pixel data of a part of asking a plurality of pixels, or ask in these any one method with the value behind the coefficient of regulation on duty etc.Dwindle processing through this, can obtain taking place on the surface original digital image data that dwindles of big variation by original digital image data (captured image data).In addition, when the mean value of the pixel data that uses a plurality of pixels, generate a pixel through using this mean value, the recovery of image that has alleviated the noise of white noise (white noise) etc. becomes possibility.In addition, handle the original digital image data that dwindles obtain, also can be used as the monitor institute images displayed data of digital camera etc. and use through dwindling.
(processing and amplifying)
Fig. 4 representes the summary of processing and amplifying.A unit of the foursquare grid among the figure, of the pixel of expression composing images data.In addition, pixel a, b, c and d are corresponding with represented pixel a, b, c and the d of Fig. 3 (C) respectively.And each pixel is the pixel with pixel data of the symbol that is affixed to each pixel.Between pixel a shown in Fig. 3 (C) and b, insert the pixel of pixel data of mean value ((a+b)/2) of the pixel data of pixel data with pixel a and pixel b.Between pixel a and the c, between pixel c and the d, between pixel b and the d, also carry out the insertion of same pixel between pixel a and the d.At this, the pixel data of the pixel at the center of inserting between pixel a and the d also can be the mean value ((b+c)/2) of pixel data of pixel data and the pixel c of pixel b.This is because the pixel at this center is the pixel that is present between pixel b and the c.According to identical viewpoint, the pixel at this center also can be the pixel of the pixel data of mean value ((a+b+c+d)/4) with pixel data of pixel a, b, c and d.Through this processing and amplifying, can return to the processing and amplifying of the pixel count that dwindles the original digital image data before handling shown in Fig. 3 (C).Fig. 4 just extracts the part of view data out and representes the situation of processing and amplifying.Be actually all pixels of view data are carried out processing and amplifying shown in Figure 4.In the signal processing apparatus 1 that this example relates to, adopted processing and amplifying shown in Figure 4.Carry out following explanation as prerequisite.
In addition, processing and amplifying is that in constituting the pixel of dwindling restored data that obtains at last through stating first reprocessing later, the new pixel data that will generate according to the pixel data that desire is amplified inserts the processing between a plurality of pixels.At this; " the new pixel data that the pixel data that amplifies according to desire generates "; Be not limited to the mean value of the pixel data of adjacent a plurality of pixels, also can utilize separation place pixel data or adjacent a plurality of pixels a part pixel data mean value or in these any one mean value multiply by data behind the coefficient of regulation etc.Through this processing and amplifying, can access with after state and dwindle original digital image data and compare exterior design big variation does not take place, but the enlarged image data that are exaggerated.
(reprocessing)
In the reprocessing that the signal processing apparatus 1 that this example relates to carries out, second reprocessing that has first reprocessing handled as the signal restoring of dwindling the view data after the processing, handles as the signal restoring of the view data after the processing and amplifying.In addition, exist as after state such as dwindle handle and the signal restoring processing of the original digital image data of processing and amplifying the 3rd repeat processing.At first, general common processing in these three reprocessings is described following.
The common processing is; The data that handling part 4 utilization changes the essential factor information are generated relatively with after the data by view data arbitrarily; The data allocations of difference that will become the view data of process object and relatively use data through the data that utilize to change essential factor information is in above-mentioned view data arbitrarily; And the generation restored data replaces above-mentioned view data arbitrarily and uses this restored data to lay equal stress on complex phase with the processing of handling.
According to Fig. 5 the summary of common processing is described.Fig. 5 is the process chart that is used for the related handling procedure of the explanation reprocessing that recovery is handled as signal (image).In Fig. 5, " I
0" be view data arbitrarily, " J " is the view data that becomes process object, these data are the data of image that are stored in the recording portion of handling part 4.At this, because " I
0" be view data arbitrarily, therefore also can use the view data " J " that becomes process object as view data arbitrarily." I
0" represent this view data I arbitrarily
0The data of modified-image, these data become be used for comparison relatively use data." G " is the variation function (some transform) of being calculated by the variation essential factor information that is stored in essential factor information storing section 7 (=deterioration essential factor information).
" δ " becomes the view data J of process object and relatively uses data I
0The data of difference." k " is based on the distribution ratio of the data that change essential factor information." I
0+n" be according to by changing variation function (some transform) G that essential factor information is calculated, the data δ of difference is allocated in view data I arbitrarily
0And the data of newly-generated restored image (restored data).At this, " h " is the destination image data of the no deterioration of finally wishing to get, and then destination image data h and the relation that becomes the view data J of process object can represent with following formula (1).
J＝h×G ……(1)
At this, " * " is the operator (below identical) of expression superposition integral.
The handling procedure of handling part 4 is, at first, and from preparing view data I arbitrarily
0Beginning (step S101).In step S102, replace destination image data h and the substitution view data I arbitrarily in (1) formula
0, obtain and relatively use data I as modified-image
0Then, relatively become the view data J of process object and relatively use data I
0, the data δ (step S103) that calculates difference.
In step S104, whether each absolute value of data δ of judging this difference if setting above, then among step S105s generates the processing of the data (=restored data) of new restored image less than setting.That is, according to changing function G, with the view data J that relatively becomes process object with relatively use data I
0The time the data δ of difference be allocated in view data I arbitrarily
0Thereby, generate new restored data I
0+1Then, repeating step S102, S103, S104 and S105.
In step S104, under the situation of absolute value of the data δ of difference, finish reprocessing less than setting.And, the restored data I when being through with reprocessing
0+nBe inferred as destination image data h.That is, during less than setting, be judged as to become and relatively use data I at the absolute value of the data δ of difference
0+nThe restored data I on basis
0+nVery be similar to destination image data h, and with this restored data I
0+nBe inferred as destination image data h.In addition, also can in recording portion 5, write down view data I arbitrarily in advance
0, change function G, be delivered to handling part 4 as required.
The words that the viewpoint of above-mentioned common processing is summarized are described below.That is, in this processing method, separating as inverse problem of process object do not solved, but solve as obtaining the optimization problem of reasonably separating.Solving as inverse problem is possible in theory, still, is difficult as realistic problem.
Under situation about solving, in the common processing that this example relates to, be prerequisite with following condition as optimization problem.
That is, (1) is defined as unique value with respect to the output of input.
(2) if the output of contrast is identical, then these inputs are identical.
(3) identical for the output that makes contrast, through when upgrading input, carrying out repeatedly
Handle and make and separate convergence.
In other words, shown in Fig. 6 (A), (B), if can generate with the view data J that becomes process object approximate relatively use data I
0(I
0+n), then become the view data I of the basic data of this generation
0(I
0+n), become the data that are similar to destination image data h.
Then; According to Fig. 7, Fig. 8, Fig. 9, Figure 10, Figure 11, Figure 12, Figure 13 and Figure 14; With with destination image data h as base image data I mg, to become the view data J of process object be example as the situation of original digital image data Img ' (afterwards state the 3rd repeat to handle), the details of the recovery (repeated treatments of step S102, S103, S104, S105) of the rolling motion video that utilizes common processing shown in Figure 5 is described.
(the recovery algorithm that rocks)
When not shaking,, in the time for exposure, concentrate on this pixel corresponding to the luminous energy of determined pixel.In addition, when shaking, luminous energy is scattered in the pixel of rocking in the time for exposure.And then, if know rocking in the time for exposure, just know the dispersing mode of the energy in the time for exposure, therefore can make unambiguous image by fuzzy image.
Below, for the sake of simplicity, describe with horizontal one dimension.With pixel begin from a left side to be made as successively S-1, S, S+1, S+2, S+3 ..., and note a certain pixel S.Because the concentration of energy in the time for exposure is in this pixel when not shaking, so the energy concentration degree is " 1.0 ".This STA representation is in Fig. 7.The shooting results of this moment is shown in the table of Fig. 8.Data shown in Figure 8 become the correct images data I mg when deterioration does not take place.In addition, each data is with the data representation of eight (bit) (0～255).
Rock in time for exposure, and the time that is respectively for 50% in the time for exposure shakes in S number pixel, time of 30% shakes in S+1 number pixel, time of 20% shakes in S+2 number pixel.The dispersing mode of energy table as shown in Figure 9 is put down in writing.This becomes by changing variation function (some transform) G that essential factor information is calculated.
Rock in all pixels identically, grasped as the problem of shift invariant amount (shift invariant).And if there is not the top to rock (vertically rocking), the situation of rocking table shown in figure 10 is put down in writing.Conduct among Figure 10 " blurred picture " and the expression data be the original digital image data Img ' of deterioration.Specifically; " 120 " of the pixel of " S-3 " for example; According to by as " 0.5 ", " 0.3 " of rocking variation function (some transform) G that change in information essential factor information calculates, the distribution ratio of " 0.2 ", in the pixel of " S-3 ", disperse to disperse in the pixel of " 60 ", " S-2 " to disperse " 24 " in the pixel of " 36 ", " S-1 ".Likewise, as " 60 " of the pixel data of " S-2 ", in " S-2 ", disperse to disperse among " 30 ", " S-1 " to disperse among " 18 ", " S " " 12 ".Variation function (some transform) G from the original digital image data Img ' and the variation essential factor information shown in Figure 9 of this deterioration are calculated calculates base image data I mg.
The I of view data arbitrarily shown in the step S101 of Fig. 5
0, when carrying out this explanation, use original digital image data Img '.That is, as I
0=Img ' begins to handle.The data that in the table of Figure 11, are used as " input " are equivalent to view data I arbitrarily
0With this data I
0, be Img ' and in step S 102, carry out superposition integral by changing variation function (some transform) G that essential factor information calculates.That is, for example, view data I arbitrarily
0" 60 " of pixel of " S-3 ", in the pixel of " S-3 ", distribute respectively in the pixel of distributing " 18 ", " S-1 " in the pixel of " 30 ", " S-2 " and distribute " 12 ".Pixel for other is distributed similarly, generates as " output I
0" and expression relatively use data I
0Therefore, the data δ of the difference of step S103 is shown in one hurdle, bottom of Figure 11.
Then, in the step S104 of Fig. 5, judge to constitute original digital image data Img ' and relatively use data I
0The absolute value of data δ of difference of a plurality of each pixel of image whether less than setting.If the result of its judgement is " denying ", then get into step S105.That is, the data of use variation essential factor information are allocated in view data I arbitrarily with the data δ of difference
0, the conduct that generates among Figure 12 " is imported " and the restored data I of expression next time
0+nAt this moment, owing to be for the first time, therefore in Figure 12, be expressed as I
0+1
The distribution of the data δ of difference is; Utilization is by changing variation function (some transform) G that essential factor information is calculated; For example will on the pixel data " 30 " of " S-3 ", multiply by the pixel that is allocated in " S-3 " as " 15 " of 0.5 gained of the distribution ratio of himself (pixel of=" S-3 "); In addition; Be distributed on the data " 15 " of the pixel of " S-2 " and multiply by as " 4.5 " that should result from 0.3 gained of the distribution ratio in this " S-2 " pixel, and then, be distributed on the data " 9.2 " of the pixel of " S-1 " and multiply by " 1.84 " that conduct should result from 0.2 gained of the distribution ratio in this " S-1 " pixel.The total amount (renewal amount) that is allocated in the pixel of " S-3 " is " 21.34 ", should be worth and view data I arbitrarily
0(using original digital image data Img ' here) addition, and generate restored data I
0+1
Shown in figure 13, this restored data I
0+1Become the data (=view data I arbitrarily of input picture of the step S102 of Fig. 5
0), execution in step S102 and to step S103 transition, thus the data δ of new difference obtained.Then, with the above-mentioned judgement of likewise carrying out step S104.If the result of its judgement is " denying ", then get into step S105, the data δ of new difference is allocated in the restored data I of last time
0+1In, generate new restored data I
0+2(with reference to Figure 14).Then, through the carrying out of step S102, by restored data I
0+2Generate the new data I of relatively using
0+2Like this, after execution in step S102, S103, if the judgement among the step S104 is " denying ", just to step S105 transition.Repeat such processing.
As above, through carrying out reprocessing, the absolute value of the data δ of difference little by little diminishes, if less than setting, and being judged as of step S104 " being " then, thus obtain rocking data that be reduced, approximate with base image data I mg.
In the restoration processing method (reprocessing of step S102, S103, S104 and S105 (the common processing)) of above-described blurred picture shown in Figure 5; The processing that utilizes handling part 4 to carry out is made up of software; But, also can be through carrying out the hardware that parts that a part handles form and constitute by sharing.In addition,, not only comprise the data of deterioration essential factor information, also comprise only making information that image changes or the information that image is improved with deterioration as the data that change essential factor information.
In addition, the number of repetition of processing also can automatically or regularly be set in signal processing apparatus 1 side.Under this situation, also can change this number of times that is set according to variation function (some transform) G that calculates by variation essential factor information.For example, also can be in the data of a certain pixel owing to rocking the number of times that increases repetition under the situation that is scattered in a plurality of pixels, and at the number of times that disperses to reduce under few situation repetition.
And then, in reprocessing, also can disperse or the energy of the view data of energy after moving does not diminish when becoming big on the contrary abort process at the data δ of difference.Whether disperse, can adopt behind the mean value of the data δ that for example observes difference if this mean value just is judged as the method for dispersing greater than the last time.In addition, in reprocessing, also can be when desire change to exceptional value with input, abort process.For example, under eight situation, the value that changes when desire is when surpassing 255 value, abort process.In addition, in reprocessing, when desire will change to exceptional value as the input of new data, also can not use this value and form normal value.For example, eight 0～255 in, will be above 255 value during in desire as the input data, handle as maximum 255.
In addition, when generation becomes the restored data of output image, there is the data conditions that exceeds the image-region of desire recovery owing to variation function (some transform) G that calculates by variation essential factor information.In this case, exceed regional data and be transfused to opposition side.In addition, existing under the data conditions that outside the zone, to import, also can be taken into this data from opposition side.For example, by be positioned at the zone below the data conditions of the data of pixel XN 1 pixel below being allocated in more under, this position is in outside the zone.Therefore, these data carry out directly over pixel XN 1, being allocated in the processing of the pixel X11 that is positioned at the top.For the adjacent pixels XN2 of pixel XN1, similarly directly over be allocated in the pixel X12 (adjacent pixels of=pixel X11) on a hurdle topmost.
In addition, in common the processing, the sampling frequency of angular velocity detection transducer is set in 60Hz～240Hz, still,, also can whenever just detects angular speed at a distance from 5 μ sec in order to detect high frequency.In addition, becoming the value of determinating reference of the data δ of difference, representing under each data conditions with eight (0～255), is " 6 " in this example.That is, less than 6 o'clock end process.In addition, through the initial data of rocking that the angular velocity detection sensor goes out, when the calibration of transducer self is insufficient, not corresponding with rocking of reality.Therefore, for corresponding, when transducer is not calibrated, must carry out the initial data that goes out through sensor multiply by the rectification of regulation multiplying power with rocking of reality.
It more than is the common processing method of handling.As based on first reprocessing of this common processing method of handling same insight in, with " the view data I arbitrarily in handling jointly
0" as the original digital image data that dwindles that the original digital image data Img ' that takes is dwindled after the processing.In addition, " relatively use data I in common the processing
0", become utilization and make this view data I arbitrarily by changing variation function (some transform) G1 that essential factor information calculates
0After (dwindling the original digital image data that dwindles after the processing) deterioration, with dwindle original digital image data be same capability relatively with dwindling data.In addition, " the becoming the view data J of process object " in common the processing becomes and dwindles original digital image data.In addition, " the data δ of difference " in common the processing becomes and dwindles original digital image data and dwindle first differential data of data with usefulness relatively.In addition, " the restored data I in common the processing
0+n", become through utilizing and first differential data is allocated in dwindles the restored data of dwindling that original digital image data generates by changing variation function (some transform) G1 that essential factor information calculates.In addition, " the destination image data h " in common the processing becomes the restored data of dwindling that obtains at last.
In second reprocessing, with " the view data I arbitrarily in handling jointly
0" conduct, will last dwindle the enlarged image data after restored data (destination image data h) is carried out processing and amplifying through what carry out that first reprocessing obtains.In addition, " relatively use data I in common the processing
0", become utilization and make this view data I arbitrarily by changing variation function (some transform) G2 that essential factor information calculates
0Relatively use amplification data after (carry out processing and amplifying after enlarged image data) deterioration.In addition, " the becoming the view data J of process object " in common the processing becomes with original digital image data Img ' or relatively uses the original image similar image data of amplification data as same capability.In addition, " the data δ of difference " in common the processing becomes the enlarged image data and relatively uses second differential data of amplification data.In addition, " the restored data I in common the processing
0+n", become through utilizing variation function (some transform) G2 that calculates by variation essential factor information that second differential data is allocated in the amplification restored data that the enlarged image data generate.In addition, " the destination image data h " in common the processing is the amplification restored data (base image data I mg or similar image data) that obtains at last.
The 3rd to repeat to handle be the common processing itself that the specified image data are carried out.That is, in the 3rd repeats to handle, " the view data I arbitrarily in common the processing
0" also can use original digital image data Img ', in addition, also can use the data of such image such as complete black, complete white, full ash, chequered with black and white tartan.In addition, " relatively use data I in common the processing
0" in the 3rd repeats to handle, become utilization and make this view data I arbitrarily by changing variation function (some transform) G3 that essential factor information calculates
0Relatively use data after (view data arbitrarily) deterioration.In addition, " the becoming the view data J of process object " in common the processing becomes original digital image data Img ' in the 3rd repeats to handle.In addition, " the data δ of difference " in common the processing becomes original digital image data Img ' and the 3rd differential data of relatively using data in the 3rd repeats to handle.In addition, " the restored data I in common the processing
0+n" in the 3rd repeats to handle, become through utilizing by changing variation function (some transform) G3 that essential factor information calculates the 3rd differential data is allocated in the restored data that view data arbitrarily generates.In addition, " the destination image data h " in common the processing becomes base image data I mg in the 3rd repeats to handle.
(whole handling process)
That Figure 15 representes is that the handling part 4 of the signal processing apparatus 1 that this example relates to mainly carries out, the flow chart of whole processing.At first, obtain among variation function (some transform) G that calculates by the variation essential factor information that obtains through test section 6, only with rock relevant information Ga, and judge degree that whether Ga below setting, promptly rock whether big (step S201).If below the setting, under the little situation of the degree of rocking, carry out the 3rd and repeat to handle (step S202), obtain the base image data at Ga, handle thereby finish image restoration as the restored image data that finally obtain.At this; Under the little situation of the degree of rocking, adopt the 3rd reason that repeat to handle to be because; Dwindle the words of processing and processing and amplifying to rocking little view data; How the original digital image data Img ' of vibrating state will dwindle or will amplify owing to will be in not, and the danger that causes deterioration of image generation on the contrary or processing time and increase becomes greatly.In addition; So only surpass under the situation of setting in the variation (being Ga) that obtains by variation essential factor information at this; Dwindle processing, first reprocessing, processing and amplifying and second reprocessing; Under the situation below the setting, carry out the 3rd processing method that repeat to handle, be not necessarily necessaryly, can not have the handling process of step S201 and S202 according to the formation such as image of the specification of signal processing apparatus 1 or processing.In addition, in step S201, also can replace with rocking relevant information Ga the information that comprises the information beyond rocking of using.
Be judged as under the situation that Ga surpasses (step S201) under the situation of setting, the degree of promptly rocking is big, the capacity of the original digital image data Img ' that at first carries out shooting is obtained dwindles dwindles processing (with reference to Fig. 3 (C)) (step S203).Then, carry out first reprocessing (step S204) to dwindling original digital image data.Then, carry out the last capacity that dwindles restored data that first reprocessing and obtain is enlarged into the processing and amplifying (with reference to Fig. 4) (step S205) identical with the capacity of original digital image data Img '.Enlarged image data after this processing and amplifying are carried out second reprocessing (step S206); And the restored image data that will finally obtain obtain as the base image data, thereby the image restoration processing that the handling part 4 of the signal processing apparatus 1 that this example relates to carries out finishes.
Under limited situation of processing time, the number of times of this second reprocessing is good with the number of times that is less than first reprocessing.Through carrying out like this, can carry out first reprocessing repeatedly, the recovering quality that dwindles restored data becomes well.In addition, meanwhile, also can be through using dwindling the auxiliary ground of restored data (confirming ground) and carrying out second reprocessing of this good quality, and the recovering quality that becomes jumbo enlarged image data through processing and amplifying is also become well.
In addition, each reprocessing of first, second and third, the precision of the image restoration along with the number of times that increases repetition (not being pro rata) improves usually.On the other hand, the whole needed processing time of image restoration, elongated pro rata with the number of times that repeats.In addition, if identical number of repetition, a needed time of reprocessing is elongated along with the capacity change of view data is many.
At this, the view data of hypothesis original digital image data Img ' is five mega pixels, dwindles the experimental example of original digital image data for equal about 300,000 pixels (minification is about 6%) of VGA (Video Graphics Array, video and graphic array of display) time and describe.In this experiment, in the processing time of carrying out 50 times first reprocessings, and original digital image data Img ' is carried out three times of repeating to handle of three order and equate.Then, carry out twice second reprocessings, its processing time equated with the times of original digital image data Img ' being carried out the triple multiple processing of two order.Will to original digital image data Img ' carried out the restored image of five order three when repeating to handle, with carried out 50 times first reprocessings after amplification restored image when carrying out twice second reprocessings again compare the obvious precision of the image restoration of the latter's amplification restored image and quality height.In addition, be equal with having carried out the restored image of 50 order three when repeating to handle.
In addition, as the additive method of the speed of accelerating reprocessing, the method that can consider to make the feedback oscillator (feedback gain) (" k " shown in Figure 5: based on the distribution ratio of the data that change essential factor information) in the reprocessing to increase.But,, then have " k δ " shown in Figure 5 thereby too become and in reprocessing, disperse the situation that can't restrain (value of δ is fully reduced) greatly if the value of feedback oscillator " k " is too increased.Therefore, make the value of feedback oscillator become big and carry out the 3rd and repeat to handle, it is all suitable generally to think.In addition, there is situation through the noise that increases white noise that number of repetition causes stressing that original digital image data Img ' is had etc.But, in this example, dwindle and amplify respectively handle the time, the noise that is comprised in the pixel data also averages out (dispersion), thus its influence is alleviated.In addition; Owing to the approaching good recovering quality of the enlarged image data that obtain by first reprocessing; Therefore the data δ of the difference in second reprocessing fully diminishes, thereby even the value change of the feedback oscillator in second reprocessing " k " is big, it is big that k δ can too not become yet.Therefore, image restoration that the signal processing apparatus 1 in this example carries out is handled, and compares with carrying out the situation that first reprocessing or the 3rd repeats to handle, and the disadvantage that in second reprocessing, causes because of the value that increases feedback oscillator is little.Therefore, in the signal processing apparatus 1 that this example relates to, can seek than carry out rapidization that the image restoration further of the 3rd situation that repeat to handle is handled.
More than, the signal processing apparatus in the example of the present invention 1 is illustrated, still, the present invention is only otherwise breaking away from its main idea just can implement various changes.For example, the restored image data that finally obtain, also can be become with the base image data dwindle or amplify after the similar image data of similarity signal data, rather than base image data.In order to obtain the similar image data, when processing and amplifying, generate the enlarged image data different (original image similar image data) with the capacity of base image.Then, original digital image data Img ' is dwindled processing or processing and amplifying, so that original digital image data Img ' is identical with original image similar image data capacity, and the adjusted original digital image data of the capacity that obtains.Then, replace view data J and the adjusted original digital image data of use capacity of the process object of the step S103 become Fig. 5, and carry out second reprocessing.
Other the method that is used to obtain the similar image data is, for example only dwindles and handles and first reprocessing, and will last dwindle the method that restored data obtains as the similar image data through what carry out that first reprocessing obtains.If this method of employing owing to can omit second reprocessing, therefore can make whole image restoration handle rapidization.When this method of employing, will handle the capacity that dwindles original digital image data that obtains through dwindling, it is identical to be adjusted into the carrying out that finally obtains the capacity that restores image data processed.In addition, the similar image data that obtain by this method, also can carry out with first reprocessing obtain at last dwindle that restored data is amplified and the processing that obtains the amplification data identical with the similar image data capacity.The method of the amplification under this situation also can be processing and amplifying (with reference to Fig. 4) method in addition that this example relates to.
Along with the exploring degree of original digital image data Img ' uprises, picture quality improves, carry out that image restoration is handled and the recovering quality of the view data that finally obtains becomes well.Therefore; The advantage that obtains the similar image data is; For example; Obtaining high-quality original digital image data Img ' through high-resolution afterwards, dwindling and handle and various reprocessing (image restoration processing) etc., can obtain finally rocking less and comparing that original digital image data Img ' capacity is little, applicability preferable image data etc.In addition, the exploring degree of the similar image data that finally obtain through signal processing apparatus 1 can be set regularly, also can be multiple setting.And, in digital camera, select or the mechanism of the exploring degree of the image that the adjustment operator finally wishes to get is good to be provided with as the signal processing apparatus that can access the similar image data 1.
The image restoration that signal processing apparatus 1 in this example carries out is handled; Shown in figure 15, carry out, finish with the order of dwindling processing (step S203), first reprocessing (step S204), processing and amplifying (step S205), second reprocessing (step S206).But, also can be for example afterwards in second reprocessing (step S206), dwindle once more handle and first reprocessing after finish.And then, also can the processing integral body with the image restoration in sequence of dwindling processing (step S203), first reprocessing (step S204), processing and amplifying (step S205), second reprocessing (step S206) shown in Figure 15 be repeated repeatedly.In addition, also can front and back that image restoration is handled, or various processing (step S203～step S206) between any period, carry out that so-called γ corrects etc., handle irrelevant processing with image restoration.
Signal processing apparatus in this example is 1 that carry out, handle with dwindling of Fig. 3 (C) expression and be, as the capacity of x, original digital image data during as y, the value (minification) that makes x/y is about 0.444 processing at the capacity that will dwindle original digital image data.The minification of dwindling processing that signal processing apparatus 1 carries out is being for good more than 0.05.Its reason is, for fear of the excessive forfeiture of the pixel data of original digital image data Img '.In addition, this minification to be being good below 0.9, with below 0.5 for better.Its reason is, in order promptly to carry out first reprocessing.
In first and second reprocessings that signal processing apparatus 1 in this example carries out; As shown in Figure 5; Dwindle original digital image data with relatively with the situation of absolute value of absolute value and the enlarged image data of the first differential data δ that dwindles data and the second differential data δ that relatively uses amplification data less than setting under (step S104), finish first reprocessing and second reprocessing.But; The condition of the end of first reprocessing; Can be with dwindling original digital image data and relatively contrast with dwindling data, become the situation below the setting according to this results of comparison or first differential data, perhaps the number of repetition of first reprocessing be the situation of setting.In addition; The condition of the end of second reprocessing; Can be with original digital image data or relatively use amplification data relatively to contrast with amplification data as the original image similar image data of same capability and above-mentioned; Become the situation below the setting according to these results of comparison or above-mentioned second differential data, the number of repetition of perhaps above-mentioned second reprocessing is the situation of setting.In addition, the judgment standard of step S201 also can be "＜", rather than "≤".
At this, above-mentioned " contrast " comprises that the corresponding relation of regulation with a plurality of data of contrast compares.For example, even seem that the view data that compare is not approximate, if the specific part of two data satisfies certain relation, the processing that handling part 4 also can make reprocessing shown in Figure 5 finish.In addition, also having the first and second differential data δ also can be the situation of the simple difference of respective pixel, still, and in general according to difference by changing variation function (some transform) G that essential factor information calculates.For example, the first and second differential data δ represent with following formula (2).
δ＝f(J、I
0、G)……(2)
In this example, will restore object as view data.But these restore viewpoint and the gimmick of handling, and can be applicable to that the recovery of all data is handled.For example, can be applicable to the recovery etc. of the voice data of numeral.Under the situation of sound signal data, can likewise handle with the pixel data in the view data making data after sound quantizes at regular intervals as the signal factor data.
In addition, above-mentioned each signal restoring method also can be by sequencing.In addition, also can be with being deposited in medium, for example CD, DVD, USB storage by the signal restoring method of sequencing, and can read through computer.Under this situation, signal processing apparatus 1 also can be downloaded the external server that is deposited in signal processing apparatus 1 by the signal restoring method of sequencing in this medium and used as required.Under this situation, signal processing apparatus 1 has the communicator that the program in this medium is downloaded.
Claims (7)
Applications Claiming Priority (3)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| JP2006349218 | 2006-12-26 | ||
| JP349218/2006 | 2006-12-26 | ||
| PCT/JP2007/074790 WO2008084670A1 (en) | 2006-12-26 | 2007-12-25 | Signal processing device |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN101569176A CN101569176A (en) | 2009-10-28 |
| CN101569176B true CN101569176B (en) | 2012-12-12 |
Family
ID=39608566
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN200780048123.0A Expired - Fee Related CN101569176B (en) | 2006-12-26 | 2007-12-25 | Signal processing device |
Country Status (3)
| Country | Link |
|---|---|
| JP (1) | JP4718618B2 (en) |
| CN (1) | CN101569176B (en) |
| WO (1) | WO2008084670A1 (en) |
Families Citing this family (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105812553A (en) * | 2015-09-24 | 2016-07-27 | 维沃移动通信有限公司 | Rapid shooting method and mobile terminal |
Family Cites Families (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2002300459A (en) * | 2001-03-30 | 2002-10-11 | Minolta Co Ltd | Image restoring device through iteration method, image restoring method and its program, and recording medium |
| JP2004343483A (en) * | 2003-05-16 | 2004-12-02 | Acutelogic Corp | Device and method for correcting camera-shake and device for detecting camera shake |
| JP4539280B2 (en) * | 2004-10-19 | 2010-09-08 | ソニー株式会社 | Image processing apparatus, image processing method, and computer program |
-
2007
- 2007-12-25 WO PCT/JP2007/074790 patent/WO2008084670A1/en active Application Filing
- 2007-12-25 JP JP2008553052A patent/JP4718618B2/en active Active
- 2007-12-25 CN CN200780048123.0A patent/CN101569176B/en not_active Expired - Fee Related
Non-Patent Citations (3)
| Title |
|---|
| JP特开2004-343483A 2004.12.02 |
| JP特开2005-117462A 2005.04.28 |
| JP特开2006-121138A 2006.05.11 |
Also Published As
| Publication number | Publication date |
|---|---|
| CN101569176A (en) | 2009-10-28 |
| JP4718618B2 (en) | 2011-07-06 |
| WO2008084670A1 (en) | 2008-07-17 |
| JPWO2008084670A1 (en) | 2010-04-30 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| EP3417606B1 (en) | A method of stabilizing a sequence of images | |
| EP1500045B1 (en) | Image rotation correction for video or photographic equipment | |
| US11477382B2 (en) | Method of stabilizing a sequence of images | |
| EP1978731A2 (en) | Image stabilizing apparatus, image pick-up apparatus and image stabilizing method | |
| CN101420612A (en) | Still image compression using moving image techniques | |
| CN101198985B (en) | Image processing apparatus and image processing method | |
| CN100358342C (en) | Image processing apparatus | |
| CN101569176B (en) | Signal processing device | |
| WO2020039747A1 (en) | Signal processing device, image-capturing device, and signal processing method | |
| WO2016152358A1 (en) | Image processing device, image processing method and image processing program | |
| US7903897B2 (en) | Image processing apparatus | |
| CN101960835B (en) | Method for generating change factor information data and signal processing device | |
| CN101305598B (en) | Image processing apparatus and method | |
| JP5940392B2 (en) | Blur correction apparatus and method | |
| CN101401417B (en) | Image processing apparatus | |
| CN101310520B (en) | Image processing device and method | |
| JP4975644B2 (en) | Image processing device | |
| JP4809190B2 (en) | Signal processing apparatus and change information generation method | |
| JP5005319B2 (en) | Signal processing apparatus and signal processing method | |
| JP3925559B2 (en) | Image processing apparatus and method, recording medium, and program | |
| JP4598623B2 (en) | Image processing device | |
| CN101960488A (en) | Method of creating data relating to change factor information and signal processor | |
| JP5007245B2 (en) | Signal processing device | |
| JP4629537B2 (en) | Image processing device | |
| JP4629622B2 (en) | Image processing device |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C14 | Grant of patent or utility model | ||
| GR01 | Patent grant | ||
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20121212 Termination date: 20171225 |
|
| CF01 | Termination of patent right due to non-payment of annual fee |