Refrigerator parts appearance inspection method and system based on machine vision
By adopting machine vision technology in refrigerator appearance detection, using binary clustering and pixel point isolation optimization filter windows, combining semantic segmentation and K-means clustering, the problem of low accuracy in detection of slight chromatic aberration and gentle structural changes is solved. Achieve more efficient and accurate appearance inspection of refrigerator parts.
Patent Information
- Application Number
- CN202411188928.7
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2024-08-28
- Publication Date
- 2024-12-03
- Estimated Expiration
- 2044-08-28
AI Technical Summary
The existing refrigerator appearance chromatic aberration detection method cannot accurately identify the chromatic aberration when dealing with slight chromatic aberration and gentle structural changes, resulting in a decrease in detection accuracy.
Using machine vision-based detection method, the filter window is optimized through binary clustering and pixel point isolation, the filter window size and edge detection strategy are dynamically adjusted, the refrigerator area is extracted in combination with the semantic segmentation algorithm, and K-means clustering and mean filtering are used. Determine whether the pixel color is within the standard color 3rd order McAdam ellipse in the XY chromaticity diagram.
It improves the accuracy of identifying subtle appearance defects of refrigerator components, enhances the robustness and efficiency of detection, reduces the computational burden, and adapts to different lighting conditions and background complexity.
Smart Images

Figure CN118710645B_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the technical field of image data processing, and in particular to a method and system for detecting the appearance of refrigerator components based on machine vision. Background Art
[0002] Appearance inspection of refrigerator parts is a key step to ensure product quality. With the development of technologies such as artificial intelligence, intelligent visual inspection equipment plays an increasingly important role in this field. At present, visual inspection equipment that can fully detect the appearance defects of parts has become a standard tool in the manufacturing industry. These devices can capture the appearance images of parts and analyze and identify defects such as surface flaws, scratches, cracks, etc. through image processing algorithms. Among them, color difference detection of refrigerator appearance is an important link to ensure product quality and brand consistency. With the intensification of market competition and the improvement of consumers' requirements for product appearance, color difference detection technology has become increasingly important in the home appliance industry. At present, the detection technology in the industry is developing towards intelligence, high efficiency and high precision.
[0003] The patent document with announcement number CN111739003B discloses a machine vision method for appearance inspection, which is used for appearance inspection of products with smooth, opaque, and consistent main color. The algorithm provided by this patent document has the characteristics of high generalization, high accuracy, and sustainable optimization, and handles two main needs: (1) detection of product color; (2) detection of product appearance. By integrating multiple algorithms, the accuracy is high; using historical data, parameters can be optimized; the SIFT image matching algorithm has its own excellent performance; high generalization, applicable to products with consistent main color.
[0004] However, a common method for detecting color differences in the appearance of refrigerators is template matching. Specifically, the part to be detected is compared with the standard color to determine whether there is color difference in the part to be detected. The simple template matching algorithm does not perform filter optimization for the noise in the image and the edges of structural changes. However, if the edge is directly extracted through the edge detection algorithm, for slight and non-serious color differences and relatively gentle structural changes, such as corners, the edges are relatively regular and close to the edge of the refrigerator, and the color change is not obvious, resulting in the grayscale value of the pixel in the image not changing dramatically, making it impossible to accurately identify, which reduces the accuracy of the filtered image in judging color differences. Summary of the invention
[0005] In order to solve the problem that the existing appearance color difference detection does not perform filter optimization on the edges of noise, color difference and structural changes in the image, and the edges of color difference and structural changes cannot be accurately identified, resulting in reduced accuracy of the filtered image in judging color difference, the present invention provides a refrigerator component appearance detection method and system based on machine vision.
[0006] In a first aspect, the present invention provides a method for detecting the appearance of refrigerator parts based on machine vision, which adopts the following technical scheme: in each component image of the refrigerator image, all pixel points within the filtering window of each pixel point are clustered into two categories; the grayscale difference between the center points of the two categories of the filtering window of the pixel point is normalized as the isolation of the pixel point; in response to any category of the filtering window of the pixel point only containing the pixel point, and the isolation is greater than a preset threshold, the filtering window of the pixel point is optimized, and the side length of the optimized filtering window is positively correlated with the isolation; in response to each category of the filtering window of the pixel point, all categories contain the neighboring pixel points of the pixel point, and the number of pixels in each category is greater than a preset number , and the isolation is greater than a preset threshold, extending to both sides along a direction perpendicular to the gradient direction of the pixel point until the angle difference between the extended pixel point and the gradient direction of the pixel point is greater than a preset angle value, and the extended pixel point and the pixel point form an edge; obtaining the side length of the filtering window after the pixel point on the edge is optimized , and are the length and width of the component image respectively, To preset the adjustment threshold, is the adjustment coefficient of the pixel point, which is positively correlated with the minimum value of the isolation, the length of the edge, and the standard deviation of the distances from all pixels on all continuous edges to each boundary; the component image is filtered according to the optimized filter window of each pixel in each component image to complete the appearance detection of refrigerator parts.
[0007] The beneficial effects are as follows: by combining binary clustering and filter window optimization based on pixel isolation, subtle appearance defects on refrigerator parts can be identified more accurately, and effective detection can be achieved even in the case of slight color differences or gentle structural changes; the dynamic adjustment strategy of the filter window size takes into account factors such as pixel isolation, edge length and consistency of distance to the refrigerator image boundary, thereby improving the adaptability to different lighting conditions and background complexity; by adaptively adjusting the filter window size and edge detection strategy, unnecessary computational burden is reduced, while ensuring detection accuracy and improving the efficiency of the overall detection process.
[0008] Furthermore, each component image of the refrigerator image is an R, G, and B component image of the refrigerator area obtained by semantic segmentation in the collected refrigerator appearance image.
[0009] The beneficial effect is that by extracting the refrigerator area of the refrigerator appearance image through the semantic segmentation algorithm and obtaining the R, G, and B component images of the refrigerator area, the accuracy of subsequent detection can be improved, and background interference can be reduced, thereby completing the appearance detection of refrigerator parts more effectively.
[0010] Furthermore, the clustering algorithm of the binary clustering is K-means clustering.
[0011] Furthermore, the filtering window of the pixel point is optimized, and the side length of the optimized filtering window satisfies the following relationship: ; In the formula, is the side length of the optimized filter window for the pixel, is the preset quantity, is the isolation of the pixel.
[0012] The beneficial effect is that by optimizing the filter window side length of the pixel points so that it is proportional to the isolation of the pixel points, edge details and slight differences can be captured more accurately, thereby improving the accuracy of appearance detection of refrigerator parts.
[0013] Furthermore, the adjustment coefficient of the pixel point satisfies the following relationship: ; In the formula, is the adjustment coefficient of the pixel, is the isolation of the pixel, is the length of the edge, represents the maximum length of all edges, is the minimum value of the standard deviation of the distances from all pixels on all continuous edges to the boundaries of the component image. It is the mean of the standard deviation of the distances from all pixels on all continuous edges to the boundaries of the refrigerator image.
[0014] The beneficial effect is that by defining the adjustment coefficient of the pixel point in relation to its isolation, edge length and consistency of the distance from the edge to the image boundary, the filter window size can be adjusted more accurately, thereby improving the accuracy of the pixel point color difference judgment.
[0015] Furthermore, it also includes: in response to the isolation of the pixel point being no greater than a preset threshold, using the filtering window of the pixel point as the optimized filtering window of the pixel point; in response to the filtering window of the pixel point, each category includes the neighboring pixel points of the pixel point, and the number of pixels in any category is less than a preset number , the filtering window of the pixel point is used as the optimized filtering window of the pixel point; the number of pixels in each category of the filtering window of the pixel point is greater than the preset number , and any category does not contain the neighboring pixels of the pixel, and the filtering window of the pixel is used as the optimized filtering window of the pixel.
[0016] Furthermore, the component images are filtered according to the optimized filtering windows of each pixel in each component image to complete the appearance detection of refrigerator parts, including: using the optimized filtering windows of each pixel in each classified image to perform mean filtering on the component images to obtain filtered component images, combining the component images to obtain a filtered refrigerator image, converting the color corresponding to each pixel in the filtered refrigerator image into an XY chromaticity diagram, and determining whether the color corresponding to each pixel is within the 3rd-order MacAdam ellipse of the standard color corresponding to the refrigerator image in the XY chromaticity diagram; if not, the area corresponding to the pixel is a color difference area, thereby completing the appearance detection of refrigerator parts based on machine vision.
[0017] The beneficial effect is that by performing mean filtering on the optimized filter window and judging whether the color of the pixel point in the XY chromaticity diagram conforms to the 3rd-order MacAdam ellipse of the standard color, the color difference area of the refrigerator parts can be effectively identified, and the accuracy and robustness of appearance detection can be improved.
[0018] In a second aspect, the present invention provides a refrigerator component appearance inspection system based on machine vision, which adopts the following technical solutions:
[0019] The refrigerator component appearance detection system based on machine vision includes: a processor and a memory, wherein the memory stores computer program instructions, and when the computer program instructions are executed by the processor, the refrigerator component appearance detection method based on machine vision is implemented.
[0020] By adopting the above technical solution, the above-mentioned refrigerator component appearance detection method based on machine vision is generated into a computer program and stored in a memory to be loaded and executed by a processor, so that a terminal device is made according to the memory and the processor for easy use.
[0021] The present invention has the following technical effects:
[0022] By detecting the edges of color differences or changing structures and adaptively adjusting the filter window size, subtle appearance defects on refrigerator parts can be identified more accurately. At the same time, unnecessary computational burden is reduced, and the efficiency of the overall inspection process is improved. Even in the case of slight color differences or gentle structural changes, effective detection can be achieved, which improves the accuracy of color difference judgment, thereby improving the accuracy and robustness of appearance inspection of refrigerator parts. BRIEF DESCRIPTION OF THE DRAWINGS
[0023] By reading the detailed description below with reference to the accompanying drawings, the above and other purposes, features and advantages of the exemplary embodiments of the present invention will become readily understood. In the accompanying drawings, several embodiments of the present invention are shown in an exemplary and non-restrictive manner, and the same or corresponding reference numerals represent the same or corresponding parts.
[0024] Figure 1 It is a flow chart of a method for detecting appearance of refrigerator parts based on machine vision in an embodiment of the present invention. DETAILED DESCRIPTION
[0025] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative work are within the scope of protection of the present invention.
[0026] It should be understood that when the terms "first", "second", etc. are used in the claims, descriptions, and drawings of the present invention, they are only used to distinguish different objects, rather than to describe a specific order. The terms "include" and "comprise" used in the description and claims of the present invention indicate the presence of the described features, wholes, steps, operations, elements, and / or components, but do not exclude the presence or addition of one or more other features, wholes, steps, operations, elements, components, and / or their collections.
[0027] The embodiment of the present invention discloses a method for detecting the appearance of refrigerator components based on machine vision, referring to Figure 1 , comprising steps S1 to S6:
[0028] It should be noted that noise, color difference and edges will cause local color changes in the image, affecting the judgment of whether there is color difference. Therefore, it is necessary to perform smoothing filtering on the image before color difference detection, and try to correct the color changes caused by noise and edges, and retain the color changes caused by color difference, so as to make the detection of color difference more accurate and reduce the misjudgment rate.
[0029] Noise, color difference, and edges will all cause color changes in an image, but they also have their own characteristics in distribution. For example, noise is distributed in an isolated and random manner in an image, and the area of color change caused by noise is generally small and distributed in a dotted manner. The area of color change caused by color difference and edges is generally large and distributed in a sheet or strip shape. Since the edges of the outer surface of a refrigerator are generally regular, straight edges, and close to the edge of the refrigerator, the color change area caused by color difference is randomly located, the shape is generally irregular, and the edge shape is unpredictable. Based on these characteristics, the filter window size of each pixel is determined to achieve the color change caused by smoothing noise and structural changes, while retaining the color change caused by color difference.
[0030] S1: In each component image of the refrigerator image, all pixels within the filter window of each pixel are clustered by binary classification.
[0031] Implementers can set the pixel filtering window according to actual implementation conditions, for example, 7×7.
[0032] Specifically, each component image of the refrigerator image is an R, G, and B component image of the refrigerator area obtained by semantic segmentation in the collected refrigerator appearance image.
[0033] The image of the refrigerator's exterior surface is obtained using the camera in the detection system, and the image is divided into the refrigerator area and the irrelevant background area using a semantic segmentation algorithm. Finally, the refrigerator area is extracted.
[0034] Specifically, the clustering algorithm of the binary clustering is K-means clustering.
[0035] When performing binary clustering on the pixels on the appearance of refrigerators, K-means clustering is particularly suitable for processing data clusters with shapes close to spheres and similar sizes because of its simplicity, efficiency, fast convergence speed and intuitive results. It helps to quickly and accurately identify subtle color differences on the appearance of refrigerators.
[0036] S2: Determine the isolation of each pixel.
[0037] The isolation of the pixel point satisfies the following relationship:
[0038] ;
[0039] In the formula, is the isolation of the pixel, It is the grayscale difference between the center points of the two categories of the filtering window of the pixel.
[0040] S3: Optimize the filter window of noise pixels.
[0041] In response to any category of the filtering window of a pixel point only including the pixel point, and the isolation is greater than a preset threshold, the filtering window of the pixel point is optimized, and the side length of the optimized filtering window is positively correlated with the isolation.
[0042] It should be noted that when the isolation is greater than a preset threshold, the pixel is regarded as a noise pixel, so the filter window of the pixel needs to be appropriately increased to more effectively remove the influence of noise.
[0043] Specifically, the filtering window of the pixel point is optimized, and the side length of the optimized filtering window satisfies the following relationship:
[0044] ;
[0045] In the formula, is the side length of the optimized filter window for the pixel, is the preset quantity, is the isolation of the pixel.
[0046] Implementers can set the preset threshold and preset number according to actual implementation conditions. For example, the preset threshold is 0.25 and the preset number is 7.
[0047] S4: Obtain the edges of color difference and structural changes in the component images.
[0048] Each category of the filter window responsive to the pixel point contains the neighboring pixels of the pixel point, and the number of pixels in each category is greater than the preset number. , and the isolation is greater than a preset threshold, extending to both sides along a direction perpendicular to the gradient direction of the pixel point until the angle difference between the extended pixel point and the gradient direction of the pixel point is greater than a preset angle value, and the extended pixel point and the pixel point form an edge.
[0049] S5: Obtain the side length of the filtering window after the pixel points on the edge are optimized.
[0050] The side length of the filtering window after pixel optimization on the edge satisfies the following relationship:
[0051] ;
[0052] In the formula, The side length of the filter window after optimizing the pixels on the edge, is the preset quantity, and are the length and width of the component image respectively, To preset the adjustment threshold, is the adjustment coefficient of the pixel, is the maximum value function, is the minimum function, is the natural logarithm function, is the natural exponential function, is the floor function.
[0053] Implementers can set a preset adjustment threshold, for example, 0.3, based on actual implementation conditions.
[0054] in, This is to prevent the filter window size from exceeding the image size. This is to prevent the filter window size from being less than 3, which is the minimum side length of the filter; when When , the color change of the pixel is caused by color difference, and the natural logarithm function is used. The filter window is reduced. The larger the value, The larger the value, the greater the reduction of the filter window, so the side length of the filter window The smaller it is, the more details of the color difference area can be preserved; when When , the color change of the pixel is caused by structural change, using the natural exponential function Increase the filter window. The larger the value, The larger the value, the greater the increase in the filter window, thus increasing the side length of the filter window. The larger the value, the better the chromatic aberration smoothing effect caused by structural changes.
[0055] Specifically, the adjustment coefficient of the pixel point satisfies the following relationship:
[0056] ;
[0057] In the formula, is the adjustment coefficient of the pixel, is the isolation of the pixel, is the length of the edge, represents the maximum length of all edges, is the minimum value of the standard deviation of the distances from all pixels on all continuous edges to the boundaries of the component image. It is the mean of the standard deviation of the distances from all pixels on all continuous edges to the boundaries of the refrigerator image.
[0058] in, The larger the value is, the longer the edge to which the pixel belongs is, which means that the edge to which the pixel belongs extends longer along the straight line direction, and the greater the possibility that the pixel is an edge with structural changes. The smaller it is, the shorter the edge length to which the pixel belongs is. Because chromatic aberration has irregular edge characteristics, there are fewer pixels extending along the straight line direction. Therefore, the shorter the length of the edge where the target pixel is located along the straight line direction, the greater the possibility that the target pixel is an edge of chromatic aberration. is the minimum value of the standard deviation of the distances from all pixels on all continuous edges to the boundaries of the component image, that is, the distance consistency. is the mean of the standard deviation of the distances from all pixels on all continuous edges to the boundaries of the refrigerator image, that is, the mean of the distance consistency. When , it means that the distance consistency between the edge to which the pixel belongs and the boundary of the component image (the boundary of the refrigerator) is weak, and the possibility that the pixel is the edge of the color difference is greater. 1. Yes was reduced, and The larger the value, the greater the reduction; similarly, when When , it means that the distance consistency between the edge where the target pixel is located and the boundary of the component image (the boundary of the refrigerator) is strong, then the possibility that the target pixel is the edge of the structure change is greater. 1. Yes was enlarged, and The smaller the value, the greater the increase.
[0059] S6: filtering the component image according to the optimized filter window of each pixel point to complete the appearance inspection of refrigerator parts.
[0060] Specifically, in response to the isolation of the pixel point being no greater than a preset threshold, the filtering window of the pixel point is used as the optimized filtering window of the pixel point; in response to the filtering window of the pixel point, each category includes the neighboring pixel points of the pixel point, and the number of pixels in any category is less than the preset number. , the filtering window of the pixel point is used as the optimized filtering window of the pixel point; the number of pixels in each category of the filtering window of the pixel point is greater than the preset number , and any category does not contain the neighboring pixels of the pixel, and the filtering window of the pixel is used as the optimized filtering window of the pixel.
[0061] Specifically, filtering the component images according to the optimized filter windows of each pixel in each component image so as to complete the appearance detection of refrigerator parts includes:
[0062] The component image is mean filtered using the optimized filter window for each pixel in each classified image to obtain a filtered component image. The component images are combined to obtain a filtered refrigerator image. The color corresponding to each pixel in the filtered refrigerator image is converted to the XY chromaticity diagram, and it is determined whether the color corresponding to each pixel is within the 3rd-order MacAdam ellipse of the standard color corresponding to the refrigerator image in the XY chromaticity diagram. If not, the area corresponding to the pixel is a color difference area, and the appearance inspection of refrigerator parts based on machine vision is completed.
[0063] It should be noted that each point on the XY chromaticity diagram represents a certain color, but if two points are too close, the human eye cannot distinguish the difference. The range in which the human eye cannot perceive color changes is called color bandwidth. The research of MacAdam et al. shows that the color bandwidth at different positions of the chromaticity diagram is not the same. If two color coordinates fall within the 1st-order MacAdam ellipse, the human eye can hardly see any difference between the two. The difference between the color corresponding to the boundary of the 3rd-order MacAdam ellipse and the center color is the color difference value that the human eye can just perceive.
[0064] An embodiment of the present invention also discloses a refrigerator component appearance detection system based on machine vision, including a processor and a memory, wherein the memory stores computer program instructions, and when the computer program instructions are executed by the processor, a refrigerator component appearance detection method based on machine vision according to the present invention is implemented.
[0065] The above system also includes other components well known to those skilled in the art, such as a communication bus and a communication interface, and their configuration and functions are known in the art, so they will not be described in detail here.
[0066] In the present invention, the aforementioned memory may be any tangible medium containing or storing a program that can be used by or in conjunction with an instruction execution system, apparatus, or device. For example, a computer-readable storage medium may be any suitable magnetic storage medium or magneto-optical storage medium, or any other medium that can be used to store the required information and can be accessed by an application, module, or both. Any such computer storage medium may be part of the device or accessible or connectable to the device.
[0067] Although this specification has shown and described a number of embodiments of the present invention, it will be apparent to those skilled in the art that such embodiments are provided by way of example only. Those skilled in the art will conceive of many modifications, changes and alternatives without departing from the ideas and spirit of the present invention. It should be understood that in the practice of the present invention, various alternatives to the embodiments of the present invention described herein may be employed.
[0068] The above are all preferred embodiments of the present invention, and are not intended to limit the protection scope of the present invention. Therefore, any equivalent changes made based on the structure, shape, and principle of the present invention should be included in the protection scope of the present invention.
Claims
1. A method for detecting the appearance of refrigerator parts based on machine vision, characterized in that: include: In each component image of the refrigerator image, all pixels within the filtering window of each pixel are clustered by binary classification to obtain two categories of the filtering window of each pixel, and the grayscale difference of the center points of the two categories of the filtering window of each pixel is normalized as the isolation of each pixel; in response to any category of the filtering window of any pixel only containing the pixel, and the isolation is greater than a preset threshold, the filtering window of the pixel is optimized, and the side length of the optimized filtering window is positively correlated with the isolation; in response to each category of the filtering window of any pixel, all the neighborhood pixels of the pixel are included, and the number of pixels in each category is greater than a preset number , and the isolation is greater than a preset threshold, extending to both sides along a direction perpendicular to the gradient direction of the pixel point until the angle difference between the extended pixel point and the gradient direction of the pixel point is greater than a preset angle value, and the extended pixel point and the pixel point form an edge; obtaining the side length of the filter window of the optimized pixel point on the edge , and are the length and width of the component images before filtering, To preset the adjustment threshold, is the adjustment coefficient of the pixel point, which is positively correlated with the minimum value of the isolation, the length of the edge, and the standard deviation of the distances from all pixels on all continuous edges to all boundaries; The component images are filtered according to the optimized filter windows of each pixel in each component image, so as to complete the appearance inspection of refrigerator parts.
2. The refrigerator component appearance detection method based on machine vision according to claim 1, characterized in that: Each component image of the refrigerator image is an R, G, and B component image of the refrigerator area obtained by semantic segmentation in the collected refrigerator appearance image.
3. The refrigerator component appearance detection method based on machine vision according to claim 1, characterized in that: The clustering algorithm for the binary clustering is K-means clustering.
4. The refrigerator component appearance detection method based on machine vision according to claim 1, characterized in that: The filtering window of the pixel point is optimized, and the side length of the optimized filtering window satisfies the following relationship: ; In the formula, is the side length of the optimized filter window for the pixel, is the preset quantity, is the isolation of the pixel.
5. The refrigerator component appearance detection method based on machine vision according to claim 1, characterized in that: The adjustment coefficient of the pixel point satisfies the following relationship: ; In the formula, is the adjustment coefficient of the pixel, is the isolation of the pixel, is the length of the edge, represents the maximum length of all edges, is the minimum value of the standard deviation of the distances from all pixels on all continuous edges to the boundaries of the component image. It is the mean of the standard deviation of the distances from all pixels on all continuous edges to the boundaries of the refrigerator image.
6. The refrigerator component appearance detection method based on machine vision according to claim 1, characterized in that: Also includes: In response to the isolation of the pixel being no greater than a preset threshold, using the filtering window of the pixel as the optimized filtering window of the pixel; Each category of the filter window responsive to a pixel point contains neighboring pixels of the pixel point, and the number of pixels in any category is less than a preset number. , the filtering window of the pixel point is used as the optimized filtering window of the pixel point; The number of pixels in each category of the filter window responsive to the pixel is greater than a preset number , and any category does not contain the neighboring pixels of the pixel, and the filtering window of the pixel is used as the optimized filtering window of the pixel.
7. The method for detecting the appearance of refrigerator components based on machine vision according to claim 6, characterized in that: The filtering of the component images according to the optimized filtering windows of each pixel in each component image so as to complete the appearance detection of the refrigerator parts includes: The component images are mean filtered using the optimized filter windows for each pixel in each component image to obtain filtered component images. The filtered component images are combined to obtain a filtered refrigerator image. The color corresponding to each pixel in the filtered refrigerator image is converted to the XY chromaticity diagram to determine whether the color corresponding to each pixel is within the 3rd-order MacAdam ellipse of the standard color corresponding to the refrigerator image in the XY chromaticity diagram. If not, the area corresponding to the pixel is a color difference area, thus completing the appearance inspection of refrigerator parts based on machine vision.
8. The appearance inspection system of refrigerator parts based on machine vision is characterized by: include: A processor and a memory, wherein the memory stores computer program instructions, and when the computer program instructions are executed by the processor, the method for detecting the appearance of refrigerator parts based on machine vision according to any one of claims 1 to 7 is implemented.
Citation Information
Patent Citations
A machine vision method for appearance inspection
CN111739003B
Sports protector appearance material defect detection method and system
CN114612384A
Cable surface defect detection method
CN115222741A
Corrosion foil production abnormity identification method
CN115330767A
Visual technology-based clock spare and accessory part quality detection method
CN117132506A
