CN113916893A - Method for detecting die-cutting product defects - Google Patents
Method for detecting die-cutting product defects Download PDFInfo
- Publication number
- CN113916893A CN113916893A CN202111152950.2A CN202111152950A CN113916893A CN 113916893 A CN113916893 A CN 113916893A CN 202111152950 A CN202111152950 A CN 202111152950A CN 113916893 A CN113916893 A CN 113916893A
- Authority
- CN
- China
- Prior art keywords
- product
- defects
- area
- die
- region
- 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
- 230000007547 defect Effects 0.000 title claims abstract description 155
- 238000000034 method Methods 0.000 title claims abstract description 45
- 230000003287 optical effect Effects 0.000 claims abstract description 26
- 230000011218 segmentation Effects 0.000 claims abstract description 15
- 238000012545 processing Methods 0.000 claims abstract description 10
- 238000012937 correction Methods 0.000 claims abstract description 4
- 239000000463 material Substances 0.000 claims description 14
- 238000000605 extraction Methods 0.000 claims description 10
- 239000011159 matrix material Substances 0.000 claims description 10
- 230000002950 deficient Effects 0.000 claims description 9
- 238000001914 filtration Methods 0.000 claims description 6
- 230000002093 peripheral effect Effects 0.000 claims description 5
- PXFBZOLANLWPMH-UHFFFAOYSA-N 16-Epiaffinine Natural products C1C(C2=CC=CC=C2N2)=C2C(=O)CC2C(=CC)CN(C)C1C2CO PXFBZOLANLWPMH-UHFFFAOYSA-N 0.000 claims description 4
- 230000009466 transformation Effects 0.000 claims description 4
- 230000002708 enhancing effect Effects 0.000 claims description 3
- 238000001514 detection method Methods 0.000 abstract description 31
- 238000004458 analytical method Methods 0.000 description 5
- 239000003292 glue Substances 0.000 description 4
- 230000000877 morphologic effect Effects 0.000 description 3
- 239000003086 colorant Substances 0.000 description 2
- 238000010586 diagram Methods 0.000 description 2
- 230000000694 effects Effects 0.000 description 2
- 238000004519 manufacturing process Methods 0.000 description 2
- 208000003464 asthenopia Diseases 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000010276 construction Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000018109 developmental process Effects 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 230000002349 favourable effect Effects 0.000 description 1
- 238000007689 inspection Methods 0.000 description 1
- 238000012795 verification Methods 0.000 description 1
- 230000000007 visual effect Effects 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N21/00—Investigating or analysing materials by the use of optical means, i.e. using sub-millimetre waves, infrared, visible or ultraviolet light
- G01N21/84—Systems specially adapted for particular applications
- G01N21/88—Investigating the presence of flaws or contamination
- G01N21/8851—Scan or image signal processing specially adapted therefor, e.g. for scan signal adjustment, for detecting different kinds of defects, for compensating for structures, markings, edges
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N21/00—Investigating or analysing materials by the use of optical means, i.e. using sub-millimetre waves, infrared, visible or ultraviolet light
- G01N21/84—Systems specially adapted for particular applications
- G01N21/88—Investigating the presence of flaws or contamination
- G01N21/8851—Scan or image signal processing specially adapted therefor, e.g. for scan signal adjustment, for detecting different kinds of defects, for compensating for structures, markings, edges
- G01N2021/8887—Scan or image signal processing specially adapted therefor, e.g. for scan signal adjustment, for detecting different kinds of defects, for compensating for structures, markings, edges based on image processing techniques
Landscapes
- Engineering & Computer Science (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Signal Processing (AREA)
- Physics & Mathematics (AREA)
- Health & Medical Sciences (AREA)
- Life Sciences & Earth Sciences (AREA)
- Chemical & Material Sciences (AREA)
- Analytical Chemistry (AREA)
- Biochemistry (AREA)
- General Health & Medical Sciences (AREA)
- General Physics & Mathematics (AREA)
- Immunology (AREA)
- Pathology (AREA)
- Investigating Materials By The Use Of Optical Means Adapted For Particular Applications (AREA)
Abstract
The invention discloses a method for detecting the defects of die-cut products, which comprises the following steps: the method comprises the steps that an optical station is used for obtaining original image information of a die-cut product, the optical station comprises an area array camera, a telecentric lens and a light source unit, the area array camera, the telecentric lens and the light source unit are arranged towards the die-cut product, and the light source unit comprises a coaxial optical remote and an annular light source which are mutually switched; carrying out coarse positioning on original image information, and carrying out gray level processing to obtain a product area; and carrying out fine positioning and correction on the product area to obtain an accurate product area. Performing region segmentation on the accurate product region according to the gray value to obtain a sub-region; and extracting defect characteristics of the sub-regions, marking the defects on the original image information, and counting the number of the defects. The invention ensures that the defect characteristics in the acquired image information can be more obvious, thereby improving the detection stability. In addition, the area to be analyzed for a particular flaw may be reduced, thereby improving detection efficiency.
Description
Technical Field
The invention relates to the field of product detection, in particular to a method for detecting die-cut product defects.
Background
With the development of 3C electronic products, various novel technologies continuously appear, and the 3C electronic product detection market is unprecedented huge. The updating, upgrading and reconstruction of products are increasingly accelerated, the detection requirements are continuously increased, and a new challenge is also provided for the detection efficiency. The quality requirements of consumers for 3C electronic products are improved, so that the quality requirements for the parts of the 3C electronic products need to be refined, the defect sizes of the products are increasingly smaller, and the requirements for the detection industry are improved.
The traditional detection is manually carried out through visual observation by a microscope. The product detection demand is large, the time consumption of manual detection can be increased, visual fatigue is generated in long working time, and careless leakage can be caused by small flaws of some products. The manual detection has strong subjectivity, different product defects and difficulty in forming a uniform judgment standard, and after long-time detection fatigue, the subjective difference can be strengthened, so that the product quality is not uniform. The manual detection efficiency is increasingly low under the condition that the product quality requirement is high, so that the detection efficiency and the detection stability are low.
Disclosure of Invention
In order to overcome the defects in the prior art, the embodiment of the invention provides a method for detecting the defects of die-cut products, which is used for solving the problems of low detection efficiency and low detection stability.
The embodiment of the application discloses: a method for detecting defects of die-cut products comprises the following steps: the method comprises the steps that an optical station is used for obtaining original image information of a die-cut product, wherein the optical station comprises an area array camera, a telecentric lens and a light source unit which are arranged towards the die-cut product, and the light source unit comprises a coaxial optical remote and an annular light source which are mutually switched; carrying out coarse positioning on the original image information, and carrying out gray level processing to obtain a product area; and carrying out fine positioning and correction on the product area to obtain an accurate product area. Performing region segmentation on the accurate product region according to the gray value to obtain a sub-region; extracting defect characteristics of the subareas, wherein the defect characteristics comprise the defects and the redundant defects of product parts, the defects of foreign matters in the round hole and the defects of broken filaments on the periphery of the product; and marking the defects on the original image information, and counting the number of the defects.
Further, in the step of "acquiring original image information of the die-cut product using an optical station, the optical station including an area-array camera, a telecentric lens and a light source unit disposed toward the die-cut product, the light source unit including a coaxial optical remote and annular light source switched with each other", the method includes: respectively acquiring product image information of the front side and the back side of the die-cut product by using two optical work stations;
further, the method for obtaining the product area by performing coarse positioning and gray processing on the original image information in the step "comprises the following steps: splitting the original image information into RGB three channels; selecting a characteristic obvious channel from the RGB three channels to carry out median filtering, thereby removing isolated image noise; performing linear enhancement on the image; and carrying out binarization on the linearly enhanced image, and extracting a product area according to the gray level features.
Further, in the step "fine positioning and correcting the product area to obtain a precise product area", the method includes: expanding the preliminary product area, and performing threshold segmentation and feature extraction to obtain a product area; calculating the average (x) of all row and column coordinates of the product area1,y1) As a hypothetical center point of the product; making a circumscribed ellipse for the product region, and taking a point (x) on the circumscribed ellipse which is farthest from the assumed central pointm,ym) (ii) a Will (x)1,y1) And (x)m,ym) The included angle between the connecting line and the marking line parallel to the extending direction of the material belt is recorded as theta1And brought into a rotation matrix, thereby making the product area positive.
Further, in step "will be with (x)m,ym) The included angle between the connecting line and the marking line parallel to the extending direction of the material belt is recorded as theta1And is brought into a rotation matrix, so that the product area is rotated right ", the method comprises the following steps: performing image opening operation on the positive product area to eliminate the interference around the product outline; calculating the average value (x) of all row-column coordinates of the product area after the opening operation and the closing operation2,y2) As the actual center point of the product; will (x)2,y2) And (x)m,ym) Is connected withThe included angle between the marking lines parallel to the extending direction of the material belt is recorded as theta2And the corrected product area is obtained by being brought into the rotation matrix; and carrying out affine transformation, opening operation and closing operation on the corrected product area to obtain an accurate product area.
Further, in the step "performing region segmentation on the accurate product region according to the gray value to obtain the sub-region", the method includes: the precise product area has different gray values according to different materials; dividing the gray value threshold into different sub-areas; and performing region filling and region feature extraction on each sub-region to select the region closest to the reality.
Further, in the step of "extracting defect characteristics of the sub-area, the defect characteristics include a missing defect and an excess defect of a product part, a foreign matter defect in a round hole and a broken filament defect at the periphery of the product", the extracting of the missing defect and the excess defect of the product part includes the following steps: making a first minimum external rectangle for the product area; judging whether the length and the width of the first minimum external rectangle are within a standard interval in a product drawing, if so, determining that the first minimum external rectangle is qualified, otherwise, determining that the first minimum external rectangle is defective and marking the first minimum external rectangle; if the product part is qualified, a second minimum circumscribed rectangle is made for the intersection part of the product part, the long side, the short side, the area and the circumscribed circle of the second minimum circumscribed rectangle are judged, and if the threshold value is exceeded, the defect is determined and marked.
Further, in the step of extracting defect characteristics of the sub-region, wherein the defect characteristics comprise the missing and redundant defects of a product part, the defect of foreign matters in the round hole and the defect of broken filaments around the product, the step of extracting the defect characteristics of the foreign matters in the round hole comprises the following steps: performing median filtering and dynamic threshold segmentation on the accurate product region to obtain a rough round hole region; making a circumscribed circle on the rough circular hole area, wherein the radius of the circumscribed circle is R1(ii) a Converting the rough circular hole area into an area outer contour, and obtaining a fitting circle by adopting an algebraic approximation method for the area outer contour, wherein the radius of the fitting circle is R2(ii) a Comparing the tolerance ranges of the circumscribed circle and the fitting circle with the tolerance ranges of the round holes, and selecting the circumscribed circle and the fitting circle which are closer to each otherThe size of the round hole is used as the actual size of the round hole; and denoising, linearly enhancing and dividing the image in the actual size of the round hole, and judging whether the foreign matter defect exists according to different gray values.
Further, in the step of "extracting defect characteristics of the sub-region, the defect characteristics include a missing part and an excess defect of a product part, a defect of a foreign matter in a hole of a round hole, and a defect of a fuzz on the periphery of the product", the extracting of the defect characteristics of the fuzz on the periphery of the product includes the following steps: performing closed operation on the product area to remove the broken filament area therein, thereby obtaining the actual area r of the product1(ii) a The actual area r of the product1Differentiating with the product area to obtain a broken filament area r 2; the hair silk region r2 is divided into a plurality of parts [ r20, r21, r22.. r2n ] isolated from each other](ii) a The actual area r of the product1Expanding and expanding the expanded product actual region r1And
intersection of [ r20, r21, r22.. r2n ] is obtained, and the broken filament area excluding the incision mark is obtained.
Further, in the step of "extracting defect characteristics of the sub-region, the defect characteristics include a missing part and an unnecessary defect of a product part, a defect of a foreign matter in a hole of a circular hole, and a defect of a broken filament around the product", the method includes: storing defect information in a defect array [ a ]1,a2,a3,a4,...,an]Wherein, the array value 1 is defective, the array value 0 is non-defective, and the array bit number represents the sorted corresponding product.
The invention has the following beneficial effects:
1. the die-cut product is polished by the coaxial light source and the annular light source in turn, so that the defect characteristics in the acquired image information can be obvious, and the detection stability is improved. In addition, the product area is subjected to area segmentation to obtain sub-areas, and then the sub-areas are removed for defect analysis, so that the area to be analyzed for specific flaws can be reduced, and the detection efficiency is improved.
2. The product image information of the front surface and the back surface of the die-cutting product is obtained through the two optical work stations, so that the defect characteristics can be obtained by detecting the product image information of the front surface and the back surface when the defects are detected subsequently, and the detection accuracy is increased.
3. And correcting the position of the product area twice, and adding an opening operation and a closing operation to eliminate the interference in the picture, so that the accurate product area image is clear and the position is correct.
4. The effect of storing the detection result of the defect is achieved, and therefore follow-up query on the detection result is facilitated.
In order to make the aforementioned and other objects, features and advantages of the invention comprehensible, preferred embodiments accompanied with figures are described in detail below.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the embodiments or the prior art will be briefly described below, it is obvious that the drawings in the following description are only some embodiments of the present invention, and for those skilled in the art, other drawings can be obtained according to the drawings without creative efforts.
FIG. 1 is a schematic view showing the construction of a detecting apparatus for die-cut products according to an embodiment of the present invention;
FIG. 2 is a schematic diagram of an optical station according to an embodiment of the present invention;
FIG. 3 is a schematic block diagram of another optical station in an embodiment of the present invention;
FIG. 4 is an image of a die cut product after the front side of the die cut product has been polished with an annular light source in accordance with an embodiment of the present invention;
fig. 5 is image information of a cut product polished by a coaxial light source in an embodiment of the present invention.
Reference numerals of the above figures: 1. an optical work station; 11. an area-array camera; 12. a telecentric lens; 13. a light source unit; 131. a coaxial light source; 132. an annular light source; 2. and (5) die cutting the product.
Detailed Description
The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
As shown in fig. 1 to 5, the method for detecting the die-cut product of the present embodiment includes the following steps:
the original image information of the die-cut product 2 is acquired by using an optical station 1, wherein the optical station 1 comprises an area array camera 11, a telecentric lens 12 and a light source unit 13 which are arranged towards the die-cut product 2, and the light source unit 13 comprises a coaxial optical far and annular light source 132 which are switched mutually. Since the die-cut product 2 itself includes a plurality of parts, the die-cut product 2 is alternately polished using the coaxial light source 131 and the ring light source 132, and various defect information can be acquired. For example, the irradiation of the coaxial light source 131 can make the defect characteristics of the multi-glue, low-glue and black-glue plugged holes more obvious, and the annular light source 132 makes the defect characteristics of the plugged holes more obvious.
And carrying out coarse positioning on the original image information, and carrying out gray processing to obtain a product area. Since the die-cut product 2 is not a single color but is formed by combining various colors, different color regions thereof display different gray values after the gray processing. The product area can be distinguished from the background area to thereby acquire the product area. The product area can be selected from the parts of the image with uniform background, high resolution, strong sharpness and more than ten gray level differences with the background.
And finely positioning and correcting the product area to obtain an accurate product area, so that subsequent defect feature extraction is facilitated.
And carrying out region segmentation on the accurate product region according to the gray value to obtain a sub-region. Because the gray values of the products after the gray processing of all colors are different, the areas of all the sub-areas are divided according to the gray values, the areas to be analyzed for specific flaws can be reduced, and the detection efficiency is improved.
And extracting defect characteristics of the sub-area, wherein the defect characteristics comprise the defects and the redundant defects of product parts, the defects of foreign matters in the round hole and the defects of broken filaments on the periphery of the product, so that the defects of the type are found out in the sub-area.
And marking the defects on the original image information, and counting the number of the defects, so that the found defects are marked and stored, and subsequent inspection and verification are facilitated.
By means of the method, the cut product 2 is polished by the coaxial light source 131 and the annular light source 132 in turn, so that defect characteristics in the acquired image information can be obvious, and the detection stability is improved. In addition, the product area is subjected to area segmentation to obtain sub-areas, and then the sub-areas are removed for defect analysis, so that the area to be analyzed for specific flaws can be reduced, and the detection efficiency is improved.
Specifically, in the step "acquiring original image information of the die-cut product 2 by using the optical station 1, the optical station 1 includes an area-array camera 11, a telecentric lens 12 and a light source unit 13 which are arranged toward the die-cut product 2, and the light source unit 13 includes a coaxial optical remote and annular light source 132" which is switched with each other, and includes:
use two optics worker stations 1 to gather respectively the product image information of 2 tow sides of cross cutting product to can acquire the product image information of 2 tow sides of cross cutting product through two optics worker stations 1, and then can detect through the product image information of two sides to the front and back when making follow-up detection defect and acquire the defect characteristic, increased the rate of accuracy that detects.
Specifically, in the step "coarsely positioning the original image information and performing gray processing to obtain a product area", the method includes:
and splitting the original image information into RGB three channels, wherein in the process, the color of the product film and the color of the black glue present different gray values.
And selecting a channel characteristic obvious channel with high contrast relative to the background from the RGB three channels to perform median filtering, thereby removing isolated image noise.
For image linear enhancement, the linear enhancement formula is as follows: g 'is g × Mult + Add, g is the gray level of the original image, and g' is the gray level of the image after linear operation, so that the image is more obvious.
And binarizing the linearly enhanced image, and extracting a product region according to the gray scale features, so that the product region is obtained, the part with obvious features can be more prominent, and the further analysis of the part is convenient to follow.
Specifically, in the step "performing fine positioning and correction on the product area to obtain an accurate product area", the method includes:
and expanding the product area, performing gray threshold segmentation and extracting the area, the long axis and the short axis of the product to obtain a clearer product area. The step of expanding may allow the product region to be enlarged for subsequent analysis thereof.
Calculating the average (x) of all row and column coordinates of the product area1,y1) As the assumed center point of the product.
Making a circumscribed ellipse for the product region, and taking a point (x) on the circumscribed ellipse which is farthest from the assumed central pointm,ym)。
Will (x)1,y1) And (x)m,ym) The included angle between the connecting line and the marking line parallel to the extending direction of the material belt is recorded as theta1The identification lines parallel to the extension direction of the material belt are arranged in the direction parallel to the extension direction of the material belt and fixed in the extension direction, and are brought into a rotation matrix T which is as follows:
thereby making the product area correct and further facilitating subsequent detection.
The die-cut products 2 may present different positions and angles due to industrial manufacturing reasons, and after roughly positioning the approximate area of the product, the precise position of the product needs to be confirmed, and the information of the precise position includes the area of each die-cut product 2 which does not contain interference, the central point position of the product, and the inclination angle of the product. The method comprises the following steps of determining the center of a product, firstly, under the condition that the product contains some flaw interference, assuming the center of the product, calculating the angle of the product, and rotating the product to a position which is favorable for using a morphological algorithm by utilizing affine transformation to obtain a corrected picture. The defects on the periphery of the product can be eliminated by a morphological algorithm. The center and angle of the product are recalculated by regions processed by morphological algorithms, the values of which are more accurate than the assumed center point and angle.
Specifically, in step "will be AND (x)m,ym) The included angle between the connecting line and the marking line parallel to the extending direction of the material belt is recorded as theta1And is brought into a rotation matrix, so that the product area is rotated right ", the method comprises the following steps:
for the positive product area, the periphery of the product area can have dirt and fuzz interference, image opening operation is carried out, and the area with smaller area at the periphery of the product can be removed, so that the periphery interference of the product outline is eliminated, and the definition of the product area is improved.
Calculating the average value (x) of all row-column coordinates of the product area after the opening operation and the closing operation2,y2) As the actual center point of the product.
Will (x)2,y2) And (x)m,ym) The included angle between the connecting line and the marking line parallel to the extending direction of the material belt is recorded as theta2And is substituted into a rotation matrix, the rotation matrix T is as follows:
and obtaining the corrected product area.
And carrying out affine transformation, opening operation and closing operation on the corrected product area to obtain an accurate product area.
That is, in this embodiment, the position of the product region is corrected twice, and the steps of opening and closing operations are added to eliminate the interference in the picture, so that the accurate image of the product region is clear and the position of the product region is correct.
Specifically, in the step "performing region segmentation on the accurate product region according to the gray value to obtain the sub-region", the method includes:
the precise product area has different gray values according to different materials.
The grey value threshold is used to divide into different sub-regions. The accurate product region can be subjected to blob analysis, the gray values of all parts of the product are different, and all parts of the product can be extracted according to threshold processing operation of the gray values, so that the region segmentation is carried out.
And performing region filling and region feature extraction on each sub-region to select the region closest to the reality. The method can be used for subdividing the areas, and for each area, the area closest to actual detection needs is selected by using methods such as area filling, area feature extraction and the like, so that interference is further eliminated, and flaws are accurately extracted.
Specifically, in the step of extracting defect characteristics of the sub-region, the defect characteristics include the missing and redundant defects of the product part, the defect of foreign matter in the hole of the round hole and the defect of the broken filament at the periphery of the product, the extraction of the missing and redundant defects of the product part includes the following steps:
making a first minimum external rectangle L for the product area;
judging whether the length and the width of the first minimum circumscribed rectangle are in a standard interval [ l ] in the product drawingmin,lmax]If the product part is not in the product part, the defect is marked, so that the defect and the redundancy of the product part are marked;
and if the product part intersection part is qualified, a second minimum circumscribed rectangle L1 is made for the product part intersection part, the long side, the short side, the area and the circumscribed circle of the second minimum circumscribed rectangle L1 are judged, and if the threshold value is exceeded, the product part intersection part is determined to be defective and marked, so that the defect and the surplus of the product part intersection part are further marked. The threshold value can be a value range of the size of the intersection part of the product part
Specifically, in the step of extracting defect characteristics of the sub-region, wherein the defect characteristics include the missing and redundant defects of a product part, the defect of foreign matters in a round hole and the defect of broken filaments around the product, the step of extracting the defect characteristics of the foreign matters in the round hole comprises the following steps:
and performing median filtering on the accurate product region to remove Gaussian noise, performing dynamic threshold segmentation on the image, and selecting a region with a higher gray value than that of the peripheral region to obtain a rough circular hole region.
Making a circumscribed circle on the rough circular hole area, wherein the radius of the circumscribed circle is R1。
Converting the rough circular hole area into an area outer contour, and obtaining a fitting circle by adopting an algebraic approximation method for the area outer contour, wherein the radius of the fitting circle is R2。
And comparing the tolerance ranges of the circumscribed circle and the fitting circle with the tolerance ranges of the round holes, and selecting the actual size of the round holes which is closer to the size of the round holes from the circumscribed circle and the fitting circle. Wherein [ R ]min,Rmax]Can be the tolerance range of the round hole on the drawing, R is the actual size of the round hole, then it is obtained through the following formula:
and denoising, linearly enhancing and dividing the image in the actual size of the round hole, and judging whether the foreign matter defect exists according to different gray values.
Specifically, in the step of extracting defect characteristics of the sub-region, the defect characteristics include missing and redundant defects of product parts, defects of foreign matters in a round hole and defects of fuzz on the periphery of a product, the extraction of the defect characteristics of the fuzz on the periphery of the product includes the following steps:
performing closed operation on the product area to remove the broken filament area therein, thereby obtaining the actual area r of the product1. I.e. the actual region r1Is obtained by subtracting the broken filament area from the product area.
The actual area r of the product1And differentiating with the product area to obtain a broken filament area r 2. I.e. the broken filament region r2 is the product region minus r1Thus obtaining the product.
The hair region r2 is divided into a plurality of portions [ r20, r21, r22.. r2n ] which are isolated from each other.
The actual area r of the product1Expanding and expanding the expanded product actual region r1And [ r20, r21, r22.. r2n]And (5) solving intersection so as to obtain a broken filament area without the edge mark. In this process, the broken filaments are connected to the product itself, so that r is the distance between the filaments1And [ r20, r21, r22.. r2n]After intersection, expanded r1May intersect each of the filaments. And a gap exists between the edge mark and the product, and the edge mark is not a product defect but is inherent in the manufacturing process, so that the edge mark can be prevented from being recognized as a broken filament and mistakenly recognized as a defect. Thereby improving the accuracy of the detection method.
Specifically, in the step of "extracting defect characteristics of the sub-region, the defect characteristics include the missing and redundant defects of the product component, the defect of foreign matter in the hole of the circular hole, and the defect of the peripheral broken filaments of the product", the method includes:
storing defect information in a defect array [ a ]1,a2,a3,a4,...,an]Wherein, the array value 1 is defective, the array value 0 is non-defective, and the array bit number represents the sorted corresponding product. In this embodiment, the products are arranged in 4 × 4 rows, the array number represents the corresponding product, and n is 16. The effect of storing the detection result of the defect is achieved, and therefore follow-up query on the detection result is facilitated.
The principle and the implementation mode of the invention are explained by applying specific embodiments in the invention, and the description of the embodiments is only used for helping to understand the method and the core idea of the invention; meanwhile, for a person skilled in the art, according to the idea of the present invention, there may be variations in the specific embodiments and the application scope, and in summary, the content of the present specification should not be construed as a limitation to the present invention.
Claims (10)
1. The method for detecting the defects of the die-cut products is characterized by comprising the following steps of:
the method comprises the steps that an optical station is used for obtaining original image information of a die-cut product, wherein the optical station comprises an area array camera, a telecentric lens and a light source unit which are arranged towards the die-cut product, and the light source unit comprises a coaxial optical remote and an annular light source which are mutually switched;
carrying out coarse positioning on the original image information, and carrying out gray level processing to obtain a product area;
and carrying out fine positioning and correction on the product area to obtain an accurate product area.
Performing region segmentation on the accurate product region according to the gray value to obtain a sub-region;
extracting defect characteristics of the subareas, wherein the defect characteristics comprise the defects and the redundant defects of product parts, the defects of foreign matters in the round hole and the defects of broken filaments on the periphery of the product;
and marking the defects on the original image information, and counting the number of the defects.
2. The method for detecting die-cut product defects according to claim 1, wherein in the step of acquiring original image information of the die-cut product using an optical station including an area-array camera, a telecentric lens and a light source unit disposed toward the die-cut product, the light source unit including coaxial optical remote and annular light sources switched with each other, comprises:
and respectively acquiring product image information of the front side and the back side of the die-cut product by using two optical work stations.
3. The method for detecting the defects of the die-cut products as claimed in claim 1, wherein in the step of coarsely positioning the original image information and performing the gray processing to obtain the product area, the method comprises the following steps:
splitting the original image information into RGB three channels;
selecting a characteristic obvious channel from the RGB three channels to carry out median filtering, thereby removing isolated image noise;
for image linear enhancement, the linear enhancement formula is as follows: g ═ g × Mult + Add;
and carrying out binarization on the linearly enhanced image, and extracting a product area according to the gray level features.
4. The method for detecting die-cut product defects according to claim 1, wherein the step of finely positioning and correcting the product area to obtain a precise product area comprises:
expanding the preliminary product area, and performing threshold segmentation and feature extraction to obtain a product area;
calculating the average (x) of all row and column coordinates of the product area1,y1) As a hypothetical center point of the product;
making a circumscribed ellipse for the product region, and taking a point (x) on the circumscribed ellipse which is farthest from the assumed central pointm,ym)；
Will (x)1,y1) And (x)m,ym) The included angle between the connecting line and the marking line parallel to the extending direction of the material belt is recorded as theta1And brought into a rotation matrix, thereby making the product area positive.
5. The method for detecting die-cut product defects according to claim 4, wherein step (x) is performed bym,ym) The included angle between the connecting line and the marking line parallel to the extending direction of the material belt is recorded as theta1And is brought into a rotation matrix, so that the product area is rotated right ", the method comprises the following steps:
performing image opening operation on the positive product area to eliminate the interference around the product outline;
calculating the average value (x) of all row-column coordinates of the product area after the opening operation and the closing operation2,y2) As the actual center point of the product;
will (x)2,y2) And (x)m,ym) The included angle between the connecting line and the marking line parallel to the extending direction of the material belt is recorded as theta2And the corrected product area is obtained by being brought into the rotation matrix;
and carrying out affine transformation, opening operation and closing operation on the corrected product area to obtain an accurate product area.
6. The method for detecting the defects of the die-cut products as claimed in claim 1, wherein in the step of dividing the precise product area into sub-areas according to the gray values, the method comprises the following steps:
the precise product area has different gray values according to different materials;
dividing the gray value threshold into different sub-areas;
and performing region filling and region feature extraction on each sub-region to select the region closest to the reality.
7. The method for detecting the defects of the die-cut products according to claim 1, wherein in the step of extracting the defect characteristics of the sub-areas, the defect characteristics comprise the defects and the redundant defects of the product parts, the defects of foreign matters in the holes of the round holes and the defects of the broken filaments at the periphery of the products, the extraction of the defects and the redundant defects of the product parts comprises the following steps:
making a first minimum external rectangle for the product area;
judging whether the length and the width of the first minimum external rectangle are within a standard interval in a product drawing, if so, determining that the first minimum external rectangle is qualified, otherwise, determining that the first minimum external rectangle is defective and marking the first minimum external rectangle;
if the product part is qualified, a second minimum circumscribed rectangle is made for the intersection part of the product part, the long side, the short side, the area and the circumscribed circle of the second minimum circumscribed rectangle are judged, and if the threshold value is exceeded, the defect is determined and marked.
8. The method for detecting the defects of the die-cut products according to claim 1, wherein in the step of extracting the defect characteristics of the sub-areas, the defect characteristics comprise the defects and the redundant defects of the product parts, the foreign body defects in the round hole holes and the defects of the peripheral broken filaments of the products, the step of extracting the foreign body defect characteristics in the round hole holes comprises the following steps:
performing median filtering and dynamic threshold segmentation on the accurate product region to obtain a rough round hole region;
making a circumscribed circle on the rough circular hole area, wherein the radius of the circumscribed circle is R1；
Converting the rough circular hole area into an area outer contour, and obtaining a fitting circle by adopting an algebraic approximation method for the area outer contour, wherein the radius of the fitting circle is R2；
Comparing the tolerance ranges of the circumscribed circle and the fitting circle with the tolerance ranges of the round holes, and selecting the circumscribed circle and the fitting circle which are closer to the size of the round holes as the actual size of the round holes;
and denoising, linearly enhancing and dividing the image in the actual size of the round hole, and judging whether the foreign matter defect exists according to different gray values.
9. The method for detecting the defects of the die-cut products according to claim 1, wherein in the step of extracting the defect characteristics of the sub-areas, the defect characteristics comprise the defects and the redundant defects of the product parts, the defects of foreign matters in the holes of the round holes and the defects of the peripheral broken filaments of the products, the step of extracting the characteristics of the peripheral broken filaments of the products comprises the following steps:
performing closed operation on the product area to remove the broken filament area therein, thereby obtaining the actual area r of the product1；
The actual area r of the product1Differentiating with the product area to obtain a broken filament area r 2;
dividing the hair region r2 into mutually isolated multiple portions [ r20, r21, r22.. r2n ];
the actual area r of the product1Expanding and expanding the expanded product actual region r1And [ r20, r21, r22.. r2n]And (5) solving intersection so as to obtain a broken filament area without the edge mark.
10. The method for detecting the defects of the die-cut products according to the claim 7, 8 or 9, wherein in the step of extracting the defect characteristics of the sub-areas, the defect characteristics comprise the defects of the parts and the redundant defects of the products, the defects of foreign matters in the holes of the round holes and the defects of the broken filaments at the periphery of the products, the method comprises the following steps:
storing defect information in a defect array [ a ]1,a2,a3,a4,...,an]Wherein, the array value 1 is defective, the array value 0 is non-defective, and the array bit number represents the sorted corresponding product.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111152950.2A CN113916893A (en) | 2021-09-29 | 2021-09-29 | Method for detecting die-cutting product defects |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111152950.2A CN113916893A (en) | 2021-09-29 | 2021-09-29 | Method for detecting die-cutting product defects |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN113916893A true CN113916893A (en) | 2022-01-11 |
Family
ID=79237174
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202111152950.2A Pending CN113916893A (en) | 2021-09-29 | 2021-09-29 | Method for detecting die-cutting product defects |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113916893A (en) |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115100209A (en) * | 2022-08-28 | 2022-09-23 | 电子科技大学 | Camera-based image quality correction method and system |
| CN115393357A (en) * | 2022-10-28 | 2022-11-25 | 菲特（天津）检测技术有限公司 | Lens surface defect detection method |
| CN118982539A (en) * | 2024-10-21 | 2024-11-19 | 张家港宏羿球墨铸铁有限公司 | A method and system for visual inspection of defects in CNC machine tool castings |
Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108648168A (en) * | 2018-03-15 | 2018-10-12 | 北京京仪仪器仪表研究总院有限公司 | IC wafer surface defects detection methods |
| CN110672617A (en) * | 2019-09-14 | 2020-01-10 | 华南理工大学 | Method for detecting defects of silk-screen area of glass cover plate of smart phone based on machine vision |
| US20200072762A1 (en) * | 2017-03-08 | 2020-03-05 | Sumco Corporation | Method of inspecting back surface of epitaxial wafer, epitaxial wafer back surface inspection apparatus, method of managing lift pin of epitaxial growth apparatus, and method of producing epitaxial wafer |
| CN111145164A (en) * | 2019-12-30 | 2020-05-12 | 上海感图网络科技有限公司 | IC chip defect detection method based on artificial intelligence |
| CN111982931A (en) * | 2020-08-27 | 2020-11-24 | 惠州高视科技有限公司 | High-precision wafer surface defect detection device and detection method thereof |
| CN112884708A (en) * | 2021-01-15 | 2021-06-01 | 深圳市悦创进科技有限公司 | Method for detecting burrs of circular injection molding piece |
| CN113034474A (en) * | 2021-03-30 | 2021-06-25 | 无锡美科微电子技术有限公司 | Test method for wafer map of OLED display |
-
2021
- 2021-09-29 CN CN202111152950.2A patent/CN113916893A/en active Pending
Patent Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20200072762A1 (en) * | 2017-03-08 | 2020-03-05 | Sumco Corporation | Method of inspecting back surface of epitaxial wafer, epitaxial wafer back surface inspection apparatus, method of managing lift pin of epitaxial growth apparatus, and method of producing epitaxial wafer |
| CN108648168A (en) * | 2018-03-15 | 2018-10-12 | 北京京仪仪器仪表研究总院有限公司 | IC wafer surface defects detection methods |
| CN110672617A (en) * | 2019-09-14 | 2020-01-10 | 华南理工大学 | Method for detecting defects of silk-screen area of glass cover plate of smart phone based on machine vision |
| CN111145164A (en) * | 2019-12-30 | 2020-05-12 | 上海感图网络科技有限公司 | IC chip defect detection method based on artificial intelligence |
| CN111982931A (en) * | 2020-08-27 | 2020-11-24 | 惠州高视科技有限公司 | High-precision wafer surface defect detection device and detection method thereof |
| CN112884708A (en) * | 2021-01-15 | 2021-06-01 | 深圳市悦创进科技有限公司 | Method for detecting burrs of circular injection molding piece |
| CN113034474A (en) * | 2021-03-30 | 2021-06-25 | 无锡美科微电子技术有限公司 | Test method for wafer map of OLED display |
Non-Patent Citations (2)
| Title |
|---|
| 付贵;李克天;李阳;: "基于机器视觉的导光板模具网点测量系统研究", 制造技术与机床, no. 10, pages 59 - 61 * |
| 颜学龙;李艳;: "共烧陶瓷多层基板表面缺陷视觉检测方法研究", 电子与封装, no. 12, pages 15 * |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115100209A (en) * | 2022-08-28 | 2022-09-23 | 电子科技大学 | Camera-based image quality correction method and system |
| CN115100209B (en) * | 2022-08-28 | 2022-11-08 | 电子科技大学 | Camera-based image quality correction method and correction system |
| CN115393357A (en) * | 2022-10-28 | 2022-11-25 | 菲特（天津）检测技术有限公司 | Lens surface defect detection method |
| CN118982539A (en) * | 2024-10-21 | 2024-11-19 | 张家港宏羿球墨铸铁有限公司 | A method and system for visual inspection of defects in CNC machine tool castings |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108460757B (en) | Mobile phone TFT-LCD screen Mura defect online automatic detection method | |
| CN113688807B (en) | Self-adaptive defect detection method, device, recognition system and storage medium | |
| CN111179225B (en) | Test paper surface texture defect detection method based on gray gradient clustering | |
| CN108776140B (en) | Machine vision-based printed matter flaw detection method and system | |
| CN113916893A (en) | Method for detecting die-cutting product defects | |
| CN109816644B (en) | An automatic detection system for bearing defects based on multi-angle light source images | |
| CN103439338B (en) | Film defects sorting technique | |
| CN114820627B (en) | Textile quality detection method based on optical means | |
| CN108596880A (en) | Weld defect feature extraction based on image procossing and welding quality analysis method | |
| CN105069790A (en) | Rapid imaging detection method for gear appearance defect | |
| US20200173933A1 (en) | Defect Inspection Method and Defect Inspection Device | |
| CN105067638A (en) | Tire fetal-membrane surface character defect detection method based on machine vision | |
| CN110108712A (en) | Multifunctional visual defect detection system | |
| CN109886960A (en) | The method of glass edge defects detection based on machine vision | |
| CN114486916A (en) | Defect detection method of mobile phone glass cover based on machine vision | |
| CN113034488A (en) | Visual detection method of ink-jet printed matter | |
| CN115035092A (en) | Image-based bottle detection method, device, equipment and storage medium | |
| JP2018096908A (en) | Inspection device and inspection method | |
| CN117830287A (en) | Circuit board on-line detection and re-judgment algorithm | |
| CN113139943A (en) | Method and system for detecting appearance defects of open circular ring workpiece and computer storage medium | |
| CN116128873A (en) | Bearing retainer detection method, device and medium based on image recognition | |
| CN114166849B (en) | Method for detecting defects of printed carbon lines and moisture-sensitive film of humidity sensor | |
| CN115866502A (en) | Microphone part surface defect online detection process | |
| CN115393290A (en) | Edge defect detection method, device and equipment | |
| CN106770372B (en) | Detection method of white glass surface defects based on microscopic vision |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |