A method and system for detecting edge concave-convex and edge fluff defects of a safety belt

Abstract

The application provides a safety belt edge concave-convex and edge fluff defect detection method and system, comprising: obtaining a safety belt two-dimensional vision photo; processing the two-dimensional vision photo to obtain a safety belt contour graph and a fluff connected domain display graph; fitting an edge straight line of the safety belt contour graph and calculating the distance of each point on the edge of the safety belt contour graph to the edge straight line, judging whether there is a concave-convex interval based on the distance; calculating the area of the fluff connected domain display graph, and judging whether there is fluff based on the area. The application is an automatic detection method based on image vision, the detection speed of which can match the production speed; the operation of which is based on pixel points, and the purpose of small target detection can be accurately achieved, the detection accuracy is high, and the false alarm rate is low; the application does not need to be labeled, time is saved, and efficiency is improved; the thresholds of the concave-convex and fluff judgment of the application can be adjusted, and different production lines and past experiences of workers can be adapted.

Description

Technical Field

[0001] The present invention relates to the field of digital image technology, and in particular, to a method and system for detecting unevenness and frayed edges of seat belts. Background Art

[0002] Due to process problems, there are some defects in the production process of seat belts, including unevenness and frayed edges of the seat belt edges, which cannot be regarded as qualified products. However, it is obviously unrealistic to manually detect the tens of thousands of meters of seat belts produced daily on the production line.

[0003] The earliest method was to use sensors for detection. However, this method has several problems. First, the accuracy is not high after actual deployment; second, the sensors are in long-term contact and friction, resulting in losses and reduced sensitivity.

[0004] With the development of deep learning, it is natural to think of using object detection for unevenness and frayed edge detection. However, the actual difficulties are as follows: First, collecting and annotating data takes a lot of time; second, the unevenness and frayed edges are very small, and in a 1200*800 image, they may only occupy twenty or thirty pixels. The academic community has always had difficulty making progress in the problem of small object detection; third, one of the assumptions of object detection is that there is similarity within a class. After actually observing the pictures, the differences within the frayed edge class are large; fourth, it is difficult for the object detection model trained based on the annotated data to make threshold adjustments according to the actual situation of each production line.

[0005] Therefore, a method for detecting unevenness and frayed edges of seat belts based on traditional vision is needed. Summary of the Invention

[0006] Aiming at the defects in the prior art, the purpose of the present invention is to provide a method and system for detecting unevenness and frayed edges of seat belts, which can detect whether there are these two defects of unevenness and frayed edges on the seat belt.

[0007] According to one aspect of the present invention, a method for detecting unevenness and frayed edges of seat belts is provided, including:

[0008] Obtaining a two-dimensional visual photo of the seat belt;

[0009] Processing the two-dimensional visual photo to obtain a seat belt contour map and a frayed edge connected domain display map;

[0010] Fitting the edge straight line of the seat belt contour map and calculating the distance from each point on the edge of the seat belt contour map to the edge straight line, and judging whether there is an uneven interval based on the distance;

[0011] Calculate the area of ​​the connected region of the hair fibers in the display map, and determine whether hair fibers exist based on the area.

[0012] Preferably, the processing of the two-dimensional visual photograph includes:

[0013] The two-dimensional visual image is sequentially subjected to grayscale conversion, grayscale inversion, binarization, and opening operations to obtain a seat belt outline image;

[0014] The two-dimensional visual image is sequentially subjected to grayscale conversion, grayscale inversion, binarization, and binary image subtraction to open the image, resulting in a hair-like connected domain display map.

[0015] Preferably, fitting the edge line of the seat belt contour map and calculating the distance from each point on the edge of the seat belt contour map to the edge line, and determining whether there is a concave or convex region based on the distance, includes:

[0016] Determine the outline of the seat belt in the seat belt outline diagram;

[0017] Extract the point sets of the left and right edges of the contour;

[0018] By fitting lines to two sets of points respectively, two straight lines can be obtained;

[0019] Calculate the distance from each point in the two point sets to the nearest line among the two lines;

[0020] The distance to each point is compared with a distance threshold, and points that are greater than the distance threshold are grouped into interval segments;

[0021] Calculate the length of the interval segment and compare it with a length threshold. Interval segments that are longer than the length threshold are considered concave-convex intervals.

[0022] Preferably, the step of calculating the area of ​​the connected region of the hair strands and determining whether hair strands exist based on the area includes:

[0023] Calculate the number of pixels, i.e. the area, of each connected region in the display graph of the hair-like connected regions;

[0024] Arrange the multiple connected components in descending order according to the number of pixels;

[0025] Connected regions with an area greater than the area threshold are called hairs.

[0026] Preferably, the distance threshold, length threshold, and area threshold can all be adjusted according to the detection requirements.

[0027] Preferably, when there is a large area of ​​unevenness in the seat belt, the fitted straight line is shifted towards the inside or outside of the seat belt.

[0028] The distance from the points in the non-convex / concave area of ​​the seat belt outline to the offset straight line is large, exceeding the set threshold distance, and the points in the non-convex / concave area form the interval segment;

[0029] If the length of the interval is greater than the set length threshold, it is determined that there is a concave-convex interval.

[0030] Preferably, when the fuzz on the seatbelt is relatively fine, the fuzz can be eliminated by opening the valve, thus enabling the detection of the type and specific information of the defect.

[0031] When the area of ​​fuzz in the seatbelt is large, it cannot be eliminated by opening the belt, and it is detected as an uneven area. According to a second aspect of the present invention, a seatbelt edge unevenness and edge fuzz defect detection system is provided, comprising:

[0032] The acquisition module acquires two-dimensional visual images of the seatbelt;

[0033] The preprocessing module processes the two-dimensional visual photograph to obtain a seatbelt outline map and a hair-like connected domain display map.

[0034] The concavity / convexity detection module fits the edge line of the seat belt contour map and calculates the distance from each point on the edge of the seat belt contour map to the edge line, and determines whether there is a concavity / convexity region based on the distance.

[0035] The hair detection module calculates the area of ​​the connected region of the hair and determines whether hair is present based on the area.

[0036] According to a third aspect of the present invention, a terminal is provided, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the program, it can be used to execute the aforementioned method for detecting unevenness and fuzziness at the edge of a seatbelt, or to run the aforementioned system for detecting unevenness and fuzziness at the edge of a seatbelt.

[0037] According to a fourth aspect of the present invention, a computer-readable storage medium is provided having a computer program stored thereon, which, when executed by a processor, can be used to perform the aforementioned method for detecting unevenness and fuzziness at the edge of a seatbelt, or to run the aforementioned system for detecting unevenness and fuzziness at the edge of a seatbelt.

[0038] Compared with the prior art, the present invention has at least one of the following beneficial effects:

[0039] The method and system for detecting edge irregularities and fuzzy edges of seat belts in this embodiment of the invention is an automatic detection method based on image vision, and its detection speed can match the production speed.

[0040] The method and system for detecting edge irregularities and fuzzy edges of seat belts in this embodiment of the invention operate based on pixels, which can accurately achieve the purpose of small target detection with high detection accuracy and low false alarm rate; it does not require labeling, saving time and improving efficiency.

[0041] The method and system for detecting edge concavity and fuzz defects of seat belts in the embodiments of the present invention can adjust the threshold for judging concavity and fuzz, so as to adapt to different production lines and workers' past experience. Attached Figure Description

[0042] Other features, objects, and advantages of the present invention will become more apparent from the following detailed description of non-limiting embodiments with reference to the accompanying drawings:

[0043] Figure 1 This is a flowchart of a method for detecting edge irregularities and edge fuzz defects in a seat belt according to an embodiment of the present invention;

[0044] Figure 2 This is a schematic diagram of the image processing procedure for concavity / convexity detection in a preferred embodiment of the present invention;

[0045] Figure 3 This is a schematic diagram of the image processing procedure for hair detection in a preferred embodiment of the present invention;

[0046] Figure 4 This is a schematic diagram of the defect detection results in a preferred embodiment of the present invention. Detailed Implementation

[0047] The present invention will now be described in detail with reference to specific embodiments. These embodiments will help those skilled in the art to further understand the present invention, but do not limit the invention in any way. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of the present invention. These all fall within the scope of protection of the present invention.

[0048] Currently, the detection of bumps and fuzz is mostly done using deep learning methods. However, this method involves a large amount of data and annotations, resulting in a large processing load and low efficiency.

[0049] To address the existing technical problems, this invention provides an embodiment of a method for detecting defects such as unevenness and fuzziness at the edge of a seat belt, the process of which can be found in [reference needed]. Figure 1 The specific process is as follows:

[0050] S100, acquire a two-dimensional visual image of the seat belt;

[0051] S200 processes the two-dimensional visual photograph of S100 to obtain the seat belt outline map and the hair filament connected domain display map;

[0052] S300, fit the edge straight line of the seat belt contour diagram of S200 and calculate the distance from the point on the edge to the straight line, and determine whether there is a concave or convex region based on the distance;

[0053] S400, calculate the area of ​​the connected region of hair in S200, and determine whether hair exists based on the area.

[0054] This embodiment is an automatic detection method based on image vision. Its detection speed can match the production speed. The thresholds for judging concavity and fuzziness in the judgment process can be adjusted to adapt to different production lines and workers' past experience.

[0055] In a preferred embodiment of the present invention, step S100 involves using a camera to capture images of the front and back of the seatbelt. A traditional visual algorithm is employed to accurately detect small targets based on pixel-level data, achieving high detection accuracy and low false alarm rate. This avoids the problem of large intra-class discrepancies in deep learning; and it eliminates the need for annotation, saving time and improving efficiency.

[0056] Through long-term research in this field, the inventors collected and statistically analyzed data on unevenness and fuzziness defects, and discovered the following characteristics of these defects:

[0057] Feature 1: When only a small portion of the seatbelt has unevenness, the local unevenness does not affect the overall fitted straight line; in this embodiment of the invention, based on this feature, a straight line is fitted using the point set of the contour, and the unevenness interval is further determined by calculating the distance based on the straight line.

[0058] Feature 2: The fibers of the seat belt are relatively fine and can be corroded away; in this embodiment of the invention, this feature is utilized to corrode the fibers during image processing, and the fibers can be obtained by subtracting the image from the binarized image.

[0059] Feature 3: The seat belt has a large area of ​​unevenness, which affects the fitting of the overall straight line;

[0060] Feature 4: The area with lint in the seat belt is relatively large, and the lint cannot be eliminated by opening and closing the belt.

[0061] Features 1 and 2 above represent the majority of cases with unevenness and fuzziness defects, while features 3 and 4 represent cases with relatively few unevenness and fuzziness defects.

[0062] Based on the above research (feature 1, feature 2), in a preferred embodiment of the present invention, S200 is implemented, as detailed in [reference needed]. Figure 2 and Figure 3 .in, Figure 2This diagram illustrates the image processing steps for concavity / convexity detection. The input image is subjected to grayscale conversion, grayscale inversion, binarization, and opening operations. (a) shows the result after grayscale conversion, (b) shows the result after grayscale inversion, (c) shows the result after binarization, and (d) shows the result after opening operations. Figure 3 This is a schematic diagram of the image processing procedure for detecting hair fibers. The input image is subjected to grayscale conversion, grayscale inversion, binarization, and binary image subtraction opening operation. Among them, (a) is the result after grayscale conversion, (b) is the result after grayscale inversion, (c) is the binary image, and (d) is the image obtained after the binary image subtraction opening operation.

[0063] In this embodiment, the opening operation is performed by first erosion and then dilation, which can eliminate small noise in the image. Subtracting the image after the opening operation (without small noise at the edges) from the binary image (with small noise at the edges) leaves only the small noise, i.e., the fuzzy texture.

[0064] Based on the above research (feature 1, feature 2), in a preferred embodiment of the present invention, S300 is implemented, and the specific process is as follows:

[0065] S301, Locate the outline of the white area after the operation is completed;

[0066] S302, extract the point sets of the left and right contours of the white outline;

[0067] S303: By fitting two points to a straight line, two straight line equations are obtained.

[0068] S304, calculate the distance from all points in two point sets to the nearest straight line;

[0069] S305: Based on the distance calculation results of S304, filter out points that are greater than the distance threshold (the set distance threshold from a point to a straight line is simply referred to as the distance threshold), and form these points into interval segments;

[0070] S306, calculate the length of the interval obtained in S305 (the maximum y-coordinate of all points in the interval minus the minimum y-coordinate). All intervals with a length greater than the length threshold (the set interval length threshold is simply referred to as the length threshold) are considered as concave and convex intervals.

[0071] Based on the above research (feature 1, feature 2), in a preferred embodiment of the present invention, S400 is implemented, and the specific process is as follows:

[0072] S401, calculate Figure 3 (d) The number of pixels in each connected component, i.e., the area;

[0073] S402, sort the connected components in descending order according to the number of pixels (area size);

[0074] S403, connected components that are greater than the area threshold (the set threshold for the number of connected component pixels is referred to as the area threshold) are considered as hairs.

[0075] In practice, during the initial calculation, the operator doesn't know the area of ​​the hair fibers, making it difficult to determine the area threshold in S403. Therefore, S402 needs to be executed to sort the data and obtain the maximum and minimum area values. Then, based on the numerical definitions of convexity / concavity and hair fibers in the project requirements, the area threshold is finally determined. After the descending sort, the top K area values ​​are compared with the threshold, which improves detection efficiency. The specific value of K is chosen based on the actual project requirements.

[0076] In actual production, the distance threshold, length threshold, and area threshold can be adjusted based on different production lines and workers' past experience to make the detected bumps and fuzz more in line with production requirements.

[0077] The above embodiment utilizes characteristics 1 and 2 of the presence of unevenness and fuzziness defects to accurately and efficiently detect these defects, and finally summarizes the two detection results as the output of the detected defects. However, this embodiment can also accurately detect defects with characteristics 3 and 4.

[0078] When the actual testing situation meets characteristic 3, that is, there are large areas of unevenness in the seat belt that affect the fitting of the overall straight line: If it is a large bulge, the fitted straight line will shift to the outside of the seat belt. In this case, the distance between the non-bulge position and the fitted straight line will be larger. If it is a large depression, the fitted straight line will shift to the inside of the seat belt. Similarly, the distance between the non-depression area and the straight line will also be larger. Here, "relatively large" means greater than the aforementioned distance threshold (the aforementioned distance threshold is set under the condition of meeting feature 1). This distance threshold is relatively small compared to the distance from the non-concave or non-convex regions to the fitted straight line. That is, the distance from the points in the non-convex or non-concave regions to the fitted straight line with deviation is greater than the set distance threshold. Then, the points in the non-convex or non-concave regions are processed as in S305 to form an interval segment. Then, according to S306, the length within this interval segment is calculated, using the same length calculation method as in feature 1, by subtracting the minimum y-coordinate value from the maximum y-coordinate value among all points in the interval segment. Since the non-convex or non-concave regions are at the beginning and end of the tested seat belt, the calculated length of this interval segment (equal to the overall length of the tested portion of the seat belt) will necessarily be greater than the length threshold. The two cases of large-area concavity and large-area convexity in feature 3 will be detected by the concavity / convexity detection process. Therefore, the concavity / convexity interval detected in this embodiment of the invention is the concavity / convexity region under any condition, covering any area of ​​features 1 and 3 mentioned above. Furthermore, when performing concave-convex region detection for features 1 and 3, the distance threshold and length threshold remain unchanged.

[0079] When the actual detection situation meets characteristic 4, i.e., the area with fuzz is relatively large and cannot be eliminated by the opening operation: In this case, performing the fuzz detection process in S400 alone will not detect the defect. However, it is worth noting that in this invention, the processing of the input image involves first performing bump detection, and then fuzz detection. Large areas of fuzz on the edge will cause edge bumps (e.g., ... Figure 4 (b) If the right side has large areas of fuzz, then these large areas of fuzz will be detected during the S300 unevenness detection process. When data that meets characteristic 2 appears, the present invention can know the type and specific information of the defect; when data that does not meet characteristic 2 appears, the present invention can detect the specific information of the defect, but does not know the specific type. Therefore, in this embodiment of the present invention, the detected fuzz is the fuzz that meets characteristic 2, and the fuzz in other cases is detected as unevenness regions.

[0080] In summary, the seat belt edge unevenness and edge fuzz defect detection method in the above embodiments can detect unevenness and fuzz defects in various areas with high accuracy and low false negative rate.

[0081] Based on the same inventive concept, in other embodiments of the present invention, a seat belt edge concavity / convexity and edge fuzz defect detection system is provided, including an acquisition module, a preprocessing module, a concavity / convexity detection module, and a fuzz detection module. The acquisition module acquires a two-dimensional visual photograph of the seat belt; the preprocessing module preprocesses the two-dimensional visual photograph to obtain a seat belt contour map and a fuzz connection region display map; the concavity / convexity detection module performs edge line fitting on the seat belt contour map and calculates the distance from each point on the edge of the seat belt contour map to the edge line to determine whether concavity / convexity intervals exist; the fuzz detection module calculates the area of ​​the fuzz connection region display map to determine whether fuzz exists.

[0082] The specific implementation techniques of each module / unit in the above examples of the present invention can be referred to the steps of the seat belt edge concavity and fuzz defect detection method in the above embodiments, and will not be repeated here.

[0083] Based on the same inventive concept, in other embodiments of the present invention, a terminal is provided, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the program, it can be used to execute the aforementioned method for detecting uneven and fuzzy edges of seat belts, or to run the aforementioned system for detecting uneven and fuzzy edges of seat belts.

[0084] Optionally, the memory is used to store programs; the memory may include volatile memory, such as random-access memory (RAM), such as static random-access memory (SRAM), double data rate synchronous dynamic random-access memory (DDR SDRAM), etc.; the memory may also include non-volatile memory, such as flash memory. The memory is used to store computer programs (such as application programs, functional modules, etc. that implement the above methods), computer instructions, etc., and the aforementioned computer programs, computer instructions, etc., can be partitioned and stored in one or more memories. Furthermore, the aforementioned computer programs, computer instructions, data, etc., can be accessed by the processor.

[0085] The aforementioned computer programs, computer instructions, etc., can be stored in partitions within one or more memory locations. Furthermore, the aforementioned computer programs, computer instructions, data, etc., can be accessed by a processor.

[0086] A processor is used to execute a computer program stored in memory to implement the various steps of the methods involved in the above embodiments. For details, please refer to the relevant descriptions in the preceding method embodiments.

[0087] The processor and memory can be separate structures or integrated structures. When the processor and memory are separate structures, they can be coupled together via a bus.

[0088] Based on the same inventive concept, in other embodiments of the present invention, a computer-readable storage medium is provided, on which a computer program is stored, which, when executed by a processor, can be used to perform the aforementioned method for detecting uneven and fuzzy edges of seat belts, or to run the aforementioned system for detecting uneven and fuzzy edges of seat belts.

[0089] Computer-readable media include computer storage media and communication media, wherein communication media include any medium that facilitates the transfer of computer programs from one place to another. Storage media can be any available medium accessible to a general-purpose or special-purpose computer. An exemplary storage medium is coupled to a processor, enabling the processor to read information from and write information to the storage medium. Of course, the storage medium can also be a component of the processor. The processor and storage medium can reside in an ASIC. Alternatively, the ASIC can reside in a user device. Of course, the processor and storage medium can also exist as separate components in a communication device.

[0090] Those skilled in the art will understand that embodiments of this application can be provided as methods, systems, or computer program products. Therefore, this application can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, this application can take the form of a computer program product embodied on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0091] This application is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to this application. It should be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart illustrations. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0092] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0093] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0094] Obviously, those skilled in the art can make various modifications and variations to this application without departing from the spirit and scope of this application. Therefore, if these modifications and variations fall within the scope of the claims of this application and their equivalents, this application also intends to include these modifications and variations. The above-mentioned preferred features can be used in any combination without conflict.

Claims

1. A method for detecting defects such as unevenness and fuzziness at the edge of a seat belt, characterized in that, include: S1, Obtain a two-dimensional visual image of the seat belt; S2, process the two-dimensional visual photograph to obtain a seat belt outline map and a hair-like connected domain display map; S2, wherein the two-dimensional visual photograph is processed to obtain a seatbelt outline map and a hair-filament connected region display map, includes: The two-dimensional visual image is sequentially subjected to grayscale conversion, grayscale inversion, binarization, and opening operations to obtain a seat belt outline image; The two-dimensional visual photograph is sequentially subjected to grayscale conversion, grayscale inversion, binarization, and binary image subtraction to open the image, resulting in a display map of the hair-like connected domains. S3, Fit the edge line of the seat belt contour map and calculate the distance from each point on the edge of the seat belt contour map to the edge line, and determine whether there is a concave or convex region based on the distance; S4, calculate the area of ​​the connected component display graph of the hair, and determine whether hair exists based on the area; Step S3 involves fitting the edge line of the seatbelt profile and calculating the distance from each point on the edge of the seatbelt profile to the edge line, and determining whether there are concave or convex regions based on the distance, including: S31, Determine the outline of the seat belt in the seat belt outline diagram; S32, extract the point sets of the left and right edges of the contour; S33: By fitting lines to two point sets respectively, two straight lines are obtained; S34, calculate the distance from each point in the two point sets to the nearest line among the two lines; When there are a few uneven areas in the seat belt, execute S35, compare the distance of each point with a distance threshold, and form an interval segment with the points that are greater than the distance threshold; When there is a large area of ​​concave and convex regions in the seat belt, the fitted straight line is shifted to the inside or outside of the seat belt. S35' is executed, and the distance from the point of the non-concave and convex region at the edge of the seat belt outline to the shifted straight line is compared with a distance threshold. Points in the non-concave and convex regions that are greater than the distance threshold are grouped into the interval segment. S36, calculate the length of the interval segment and compare it with the length threshold. Interval segments that are greater than the length threshold are considered concave-convex intervals.

2. The method for detecting edge irregularities and fuzzy edges of a seatbelt according to claim 1, characterized in that, The calculation of the area of ​​the connected component display graph of the hair strand, and the determination of whether hair strands exist based on the area, includes: Calculate the number of pixels, i.e. the area, of each connected region in the graph of the hair-like connected regions; Arrange the multiple connected components in descending order according to the number of pixels; Connected regions with an area greater than the area threshold are called hairs.

3. The method for detecting edge irregularities and edge fuzz defects in a seatbelt according to claim 2, characterized in that, The distance threshold, length threshold, and area threshold can all be adjusted according to the detection requirements.

4. A system for implementing the method for detecting edge irregularities and edge fuzz defects of seat belts as described in claim 1, characterized in that, include: The acquisition module acquires two-dimensional visual images of the seatbelt; The preprocessing module processes the two-dimensional visual photograph to obtain a seatbelt outline map and a hair-like connected domain display map. The concavity / convexity detection module fits the edge line of the seat belt contour map and calculates the distance from each point on the edge of the seat belt contour map to the edge line, and determines whether there is a concavity / convexity region based on the distance. The hair detection module calculates the area of ​​the connected region display graph of the hair and determines whether hair is present based on the area.

5. A terminal, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the program, it can be used to perform the method for detecting edge irregularities and edge fuzz defects of seat belts as described in any one of claims 1-3.

6. A computer-readable storage medium having a computer program stored thereon, characterized in that, When executed by a processor, this program can be used to perform the method for detecting edge irregularities and edge fuzz defects of any one of claims 1-3.

Images