A texture-based defect detection method, system, device, and medium

Abstract

The application relates to a texture-based defect detection method, system, device and medium, and the technical scheme points are as follows: the method comprises the following steps: acquiring image information of a cloth surface; performing guided filtering processing on the image information to obtain a filtered image; extracting texture edge information of the filtered image to obtain a texture edge information image; performing connected domain processing on the texture edge information image to obtain a connected domain set; scanning each connected domain in the connected domain set to obtain a corresponding edge contour area; and performing rectangular frame labeling on each edge contour area to obtain a corresponding defect area; and the application has the advantages that the cloth texture can be filtered to quickly locate defects.

Description

Technical Field

[0001] The present invention relates to the field of image recognition technology, and more particularly to a texture-based defect detection method, system, device, and medium. Background Art

[0002] During the production process of grey fabric, except for jacquard-type grey fabric, the normal texture of the grey fabric surface appears periodically. In the process of using deep learning algorithms to detect grey fabric defects, a large number of defect samples need to be collected, and it takes a long time to collect enough defect sample data. In the actual grey fabric production process, the grey fabric is switched frequently, so the fabric types change relatively quickly. In this scenario, enough defect samples need to be collected for each changed fabric type before deep learning can be used for model training. This defect detection solution has great limitations in this case, so there is still room for improvement. Summary of the Invention

[0003] In view of the shortcomings of the existing technology, the purpose of the present invention is to provide a texture-based defect detection method, system, equipment, and medium, which have the advantage of being able to filter the texture of grey cloth to quickly locate defects.

[0004] The above technical objectives of the present invention are achieved through the following technical solutions: A texture-based defect detection method comprising:

[0005] Acquire image information of the grey fabric surface;

[0006] performing guided filtering on the image information to obtain a filtered image;

[0007] Extracting texture edge information of the filtered image to obtain a texture edge information image;

[0008] Performing connected domain processing on the texture edge information image to obtain a connected domain set;

[0009] Scanning each connected domain in the connected domain set to obtain a corresponding edge contour area;

[0010] Mark each edge contour area with a rectangular frame to obtain the corresponding defect area.

[0011] Optionally, performing guided filtering on the image information to obtain a filtered image includes:

[0012] Obtaining a surface regular texture of the image information to obtain a guide picture;

[0013] The image information is subjected to guided filtering based on the guided image to obtain a filtered image.

[0014] Optionally, performing guided filtering on the image information based on the guided picture to obtain a filtered image includes:

[0015] Input the image information, guide image, window radius r and regularization term ε into the guided filtering algorithm;

[0016] mean I =f mean (I), mean p =f mean (p), corr I =f corr (I.*I),

[0017] corr Ip =f corr (I.*p);

[0018] var I =corr I -mean I .*mean I , cov Ip =corr Ip -mean I .*mean p ;

[0019] a=cov Ip . / (var I +ε), b=mean p -a.*mean I ;

[0020] mean a =f mean (a),mean b =f mean (b);

[0021] q=mean a .*I+mean b ;

[0022] Among them, f mean It is a mean filter with a time complexity of O(N); p is the image information, I is the guide image, and q is the output filtered image.

[0023] Optionally, extracting texture edge information of the filtered image to obtain a texture edge information image includes:

[0024] Preset the variance σ1, σ2 and scaling ratios k1, k2 of the Gaussian convolution kernel to obtain the first Gaussian filter and the second Gaussian filter

[0025] Apply a first Gaussian filter to the filtered image Perform convolution to obtain a first convolution image;

[0026] Apply a second Gaussian filter to the filtered image Perform convolution to obtain a second convolution image;

[0027] Subtracting the first convolution image from the second convolution image to obtain a Gaussian difference image;

[0028] The pixel points that are 0 in the Gaussian difference image are taken to obtain a texture edge information image corresponding to the filtered image.

[0029] Optionally, performing connected domain processing on the texture edge information image to obtain a connected domain set includes:

[0030] Performing expansion processing on the texture edge information image to obtain an expanded texture image;

[0031] performing corrosion processing on the expanded texture image to obtain a corroded texture image;

[0032] All connected domains in the eroded texture image are found to obtain a connected domain set.

[0033] Optionally, scanning each connected domain in the connected domain set to obtain a corresponding edge contour area includes:

[0034] Filtering all connected domains in the connected domain set, removing connected domains whose areas are smaller than a preset area, and obtaining a standard connected domain set;

[0035] Contour scanning is performed on all connected domains in the standard connected domain set to obtain edge contour areas of opposite-sided edges corresponding to each connected domain in the standard connected domain set.

[0036] Optionally, the step of marking each edge contour area with a rectangular frame to obtain a corresponding defect area includes:

[0037] The corresponding minimum circumscribed slope distance shape is determined according to the polygonal edge of each edge contour area, and each minimum circumscribed slope distance shape is used as the defect area corresponding to each edge contour area.

[0038] A texture-based defect detection system includes: an image acquisition module for acquiring image information of a grey cloth surface;

[0039] An image filtering module, configured to perform guided filtering on the image information to obtain a filtered image;

[0040] A texture extraction module, configured to extract texture edge information of the filtered image to obtain a texture edge information image;

[0041] A connected domain identification module is used to perform connected domain processing on the texture edge information image to obtain a connected domain set;

[0042] An edge recognition module is used to scan each connected domain in the connected domain set to obtain a corresponding edge contour area;

[0043] The defect marking module is used to mark each edge contour area with a rectangular frame to obtain the corresponding defect area.

[0044] A computer device includes a memory and a processor, wherein the memory stores a computer program and the processor implements the steps of the above method when executing the computer program.

[0045] A computer-readable storage medium stores a computer program, which implements the steps of the above method when executed by a processor.

[0046] In summary, the present invention has the following beneficial effects: by performing guided filtering processing on the image information of the obtained grey fabric surface and using the texture of the grey fabric as a guide image, the interference of the grey fabric surface texture in the image information can be removed, so that the filtered image does not contain the texture of the grey fabric surface, thereby leaving texture edge information different from the grey fabric surface texture, and obtaining a texture edge information image, wherein the texture edge information is the location of the defect; subsequently, by performing connected domain processing on the texture edge information image, all connected regions in the texture edge information image are obtained, and a connected domain set is established, each connected domain must contain a corresponding defect; at this time, the edge contour area of ​​each connected domain is obtained to obtain the edge line of each connected domain, and the edge contour area corresponding to each connected domain is marked with a rectangular frame using a minimum circumscribed slant distance shape, and the minimum circumscribed slant distance shape can contain the corresponding defect; because the image information is filtered using the grey fabric texture as a guide before defect recognition, the interference of the grey fabric texture on defect recognition can be eliminated, thereby quickly locating the defect area, so that it is possible to quickly adapt defect detection for images with different surface texture results without relying on a large amount of data. BRIEF DESCRIPTION OF THE DRAWINGS

[0047] Figure 1 It is a schematic diagram of the process of the present invention;

[0048] Figure 2 This is a structural diagram of the present invention when assembled;

[0049] Figure 3 2 is a diagram showing the internal structure of a computer device in an embodiment of the present invention. DETAILED DESCRIPTION

[0050] To make the objectives, features, and advantages of the present invention more readily apparent, the following detailed description of the present invention is provided with reference to the accompanying drawings. The accompanying drawings illustrate several embodiments of the present invention. However, the present invention may be implemented in many different forms and is not limited to the embodiments described herein.

[0051] In the present invention, unless otherwise expressly specified and limited, the terms "installed", "connected", "connected", "fixed" and the like should be understood in a broad sense. For example, it can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection or an electrical connection; it can be a direct connection, or an indirect connection through an intermediate medium, or it can be a communication between the two components. For those of ordinary skill in the art, the specific meanings of the above terms in the present invention can be understood according to the specific circumstances. The terms "first" and "second" are used for descriptive purposes only and cannot be understood as indicating or implying relative importance or implicitly indicating the number of the indicated technical features. Therefore, the features defined as "first" and "second" may explicitly or implicitly include one or more of such features.

[0052] In the present invention, unless otherwise expressly specified and limited, a first feature being "above" or "below" a second feature may include the first and second features being in direct contact, or may include the first and second features being in contact not directly but through another feature between them. Moreover, a first feature being "above," "above," and "above" a second feature includes the first feature being directly above and obliquely above the second feature, or simply indicates that the first feature is higher in level than the second feature. A first feature being "below," "below," and "below" a second feature includes the first feature being directly below and obliquely below the second feature, or simply indicates that the first feature is lower in level than the second feature. The terms "vertical," "horizontal," "left," "right," "above," "below," and similar expressions are for illustrative purposes only and do not indicate or imply that the device or element referred to must have a specific orientation, be constructed and operate in a specific orientation, and therefore should not be understood as limiting the present invention.

[0053] The present invention is described in detail below with reference to the accompanying drawings and embodiments.

[0054] The present invention provides a texture-based defect detection method, such as Figure 1 Shown, including:

[0055] Step 100: Acquire image information of the surface of the grey cloth;

[0056] Step 200: performing guided filtering on the image information to obtain a filtered image;

[0057] Step 300: extracting texture edge information of the filtered image to obtain a texture edge information image;

[0058] Step 400: Perform connected domain processing on the texture edge information image to obtain a connected domain set;

[0059] Step 500: Scan each connected domain in the connected domain set to obtain a corresponding edge contour area;

[0060] Step 600: Mark each edge contour area with a rectangular frame to obtain the corresponding defect area.

[0061] In practical applications, by performing guided filtering on the image information of the obtained grey fabric surface and using the texture of the grey fabric as a guide image, the interference of the grey fabric surface texture in the image information can be removed, so that the filtered image does not contain the texture of the grey fabric surface, thereby leaving texture edge information different from the grey fabric surface texture and obtaining a texture edge information image, wherein the texture edge information is the location of the defect. Subsequently, by performing connected domain processing on the texture edge information image, all connected regions in the texture edge information image are obtained, and a connected domain set is established. Each connected domain must contain a corresponding defect. At this time, the edge contour area of ​​each connected domain is obtained to obtain the edge line of each connected domain. The edge contour area corresponding to each connected domain is then marked with a rectangular frame using a minimum circumscribed slant distance shape. The minimum circumscribed slant distance shape can contain the corresponding defect. Since the image information is filtered using the grey fabric texture as a guide before defect recognition, the interference of the grey fabric texture on defect recognition can be eliminated, thereby quickly locating the defect area. Therefore, it is possible to quickly adapt defect detection for images with different surface texture results without relying on a large amount of data.

[0062] Furthermore, performing guided filtering on the image information to obtain a filtered image includes:

[0063] Obtaining a surface regular texture of the image information to obtain a guide picture;

[0064] The image information is subjected to guided filtering based on the guided image to obtain a filtered image.

[0065] In practical applications, the texture clusters on the normal surface of the grey fabric will appear periodically, while the defects on the grey fabric will destroy the regular texture of the surface. By using the image processing filtering algorithm to filter the textures of different frequency bands, a filtered image with texture interference removed can be obtained.

[0066] Optionally, performing guided filtering on the image information based on the guided picture to obtain a filtered image includes:

[0067] Input the image information, guide image, window radius r and regularization term ε into the guided filtering algorithm;

[0068] mean I =f mean (I), mean p =f mean (p), corr I =f corr (I.*I),

[0069] corr Ip =f corr (I.*p);

[0070] var I =corr I -mean I .*mean I , cov Ip =corr Ip -mean I .*mean p ;

[0071] a=cov Ip . / (var I +ε), b=mean p -a.*mean I ;

[0072] mean a =f mean (a),mean b =f mean (b);

[0073] q=mean a .*I+mean b ;

[0074] Among them, f mean It is a mean filter with a time complexity of O(N); p is the image information, I is the guide image, and q is the output filtered image.

[0075] In practical applications, the image information p, the guide image I, and the pre-set window radius r and regularization term ε are input into the guided filtering algorithm, and the output can filter out the surface texture of the grey fabric, leaving a filtered image that is different from the surface texture of the grey fabric; among them, the accuracy of the filtering can be changed by modifying the window radius r, and the regularization term ε is used to define the smooth area and the edge area.

[0076] Optionally, extracting texture edge information of the filtered image to obtain a texture edge information image includes:

[0077] Preset the variance σ1, σ2 and scaling ratios k1, k2 of the Gaussian convolution kernel to obtain the first Gaussian filter and the second Gaussian filter

[0078] Apply a first Gaussian filter to the filtered image Perform convolution to obtain a first convolution image;

[0079] Apply a second Gaussian filter to the filtered image Perform convolution to obtain a second convolution image;

[0080] Subtracting the first convolution image from the second convolution image to obtain a Gaussian difference image;

[0081] The pixel points that are 0 in the Gaussian difference image are taken to obtain a texture edge information image corresponding to the filtered image.

[0082] In practical applications, by filtering the same filtered image with Gaussian filters with different variances and scaling ratios on both sides, and subtracting the first convolution image from the second convolution image, the edge feature points in the filtered image can be obtained. The pixel points that are 0 in the Gaussian difference image are used as feature points, indicating the parts that are completely consistent in the first convolution image and the second convolution image, thereby obtaining the texture edge information image corresponding to the filtered image.

[0083] Furthermore, the performing connected domain processing on the texture edge information image to obtain a connected domain set includes:

[0084] Performing expansion processing on the texture edge information image to obtain an expanded texture image;

[0085] performing corrosion processing on the expanded texture image to obtain a corroded texture image;

[0086] All connected domains in the eroded texture image are found to obtain a connected domain set.

[0087] In practical applications, since most parts of the texture edge information image are completely consistent, there are a large number of interconnected areas in the texture edge information image, and the boundary area in the texture edge information image is the area where the defect is located. Therefore, it is necessary to first expand the connected areas in the texture edge information image, and connect the smaller areas in the middle of the two adjacent areas to form a larger area; then perform corrosion first; this can not only remove the influence of the area expansion caused by the expansion of isolated noise points, but also remove the breaks in the small areas with weak edge sections, and find all the connected domains in the eroded texture image to obtain a connected domain set. At this time, the boundaries of each connected domain can constitute the edge contour of the defect.

[0088] Furthermore, scanning each connected domain in the connected domain set to obtain a corresponding edge contour area includes:

[0089] Filtering all connected domains in the connected domain set, removing connected domains whose areas are smaller than a preset area, and obtaining a standard connected domain set;

[0090] Contour scanning is performed on all connected domains in the standard connected domain set to obtain edge contour areas of opposite-sided edges corresponding to each connected domain in the standard connected domain set.

[0091] In practical applications, it is necessary to use the maximum connected domain algorithm to filter out connected domains that are smaller than a preset area (10 pixels). These connected domains are considered to be noise edge areas, and the contour scanning algorithm is used to find the edge contour area for the mask area of ​​the connected domain that passes a certain threshold.

[0092] Furthermore, the rectangular frame marking of each edge contour area to obtain the corresponding defect area includes:

[0093] The corresponding minimum circumscribed slope distance shape is determined according to the polygonal edge of each edge contour area, and each minimum circumscribed slope distance shape is used as the defect area corresponding to each edge contour area.

[0094] In practical applications, the size of each edge contour area is determined by the minimum circumscribed slope distance shape. There is a corresponding defect in each minimum circumscribed slope distance shape, thereby obtaining a set of all defects in the image information.

[0095] Specifically, this application does not require the collection of a large number of defect samples or model training for defect detection on the surface of grey fabrics. It only requires adjusting initialization parameters for different texture data to quickly adapt to defect detection for new fabric types. This allows for adaptation to defect detection for new fabric types in a relatively short period of time. Guided filtering can better achieve the goal of maintaining defects beyond the initial periodic texture, with good adaptability to periodic textures. At the same time, the algorithm is optimized specifically for the characteristics of the grey fabric surface and the algorithm flow.

[0096] like Figure 2 As shown, the present invention also provides a texture-based defect detection system, comprising:

[0097] An image acquisition module 10 is used to acquire image information of the surface of the grey cloth;

[0098] An image filtering module 20 is configured to perform guided filtering on the image information to obtain a filtered image;

[0099] The texture extraction module 30 is used to extract texture edge information of the filtered image to obtain a texture edge information image;

[0100] A connected domain identification module 40 is configured to perform connected domain processing on the texture edge information image to obtain a connected domain set;

[0101] An edge recognition module 50 is configured to scan each connected domain in the connected domain set to obtain a corresponding edge contour area;

[0102] The defect marking module 60 is used to mark each edge contour area with a rectangular frame to obtain the corresponding defect area.

[0103] The specific definition of a texture-based defect detection system can be found in the definition of a texture-based defect detection method above and will not be repeated here. Each module in the above-mentioned texture-based defect detection system can be implemented in whole or in part through software, hardware, or a combination thereof. Each of the above-mentioned modules can be embedded in or independent of a processor in a computer device in hardware form, or can be stored in a memory in a computer device in software form, so that the processor can call and execute the corresponding operations of each of the above modules.

[0104] In one embodiment, a computer device is provided. The computer device may be a server, and its internal structure diagram may be as follows: Figure 3 As shown, the computer device includes a processor, memory, a network interface, and a database connected via a system bus. The processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and internal memory. The non-volatile storage medium stores an operating system, a computer program, and a database. The internal memory provides an environment for the operating system and computer program in the non-volatile storage medium to run. When executed by the processor, the computer program implements a texture-based defect detection method.

[0105] Those skilled in the art will understand that Figure 3 The structure shown in the figure is only a block diagram of a part of the structure related to the solution of the present application, and does not constitute a limitation on the computer device to which the solution of the present application is applied. The specific computer device may include more or fewer components than shown in the figure, or combine certain components, or have a different component arrangement.

[0106] In one embodiment, a computer device is provided, including a memory and a processor, wherein a computer program is stored in the memory, and when the processor executes the computer program, the following steps are implemented: acquiring image information of a surface of a grey fabric;

[0107] performing guided filtering on the image information to obtain a filtered image;

[0108] Extracting texture edge information of the filtered image to obtain a texture edge information image;

[0109] Performing connected domain processing on the texture edge information image to obtain a connected domain set;

[0110] Scanning each connected domain in the connected domain set to obtain a corresponding edge contour area;

[0111] Mark each edge contour area with a rectangular frame to obtain the corresponding defect area.

[0112] In one embodiment, performing guided filtering on the image information to obtain a filtered image includes:

[0113] Obtaining a surface regular texture of the image information to obtain a guide picture;

[0114] The image information is subjected to guided filtering based on the guided image to obtain a filtered image.

[0115] In one embodiment, performing guided filtering on the image information based on the guided picture to obtain a filtered image includes:

[0116] Input the image information, guide image, window radius r and regularization term ε into the guided filtering algorithm;

[0117] mean I =f mean (I), mean p =f mean (p), corr I =f corr (I.*I),

[0118] corr Ip =f corr (I.*p);

[0119] var I =corr I -mean I .*mean I , cov Ip =corr Ip -mean I .*mean p ;

[0120] a=cov Ip . / (var I +ε), b=mean p -a.*mean I ;

[0121] meana =f mean (a),mean b =f mean (b);

[0122] q=mean a .*I+mean b ;

[0123] Among them, f mean It is a mean filter with a time complexity of O(N); p is the image information, I is the guide image, and q is the output filtered image.

[0124] In one embodiment, extracting texture edge information of the filtered image to obtain a texture edge information image includes:

[0125] Preset the variance σ1, σ2 and scaling ratios k1, k2 of the Gaussian convolution kernel to obtain the first Gaussian filter and the second Gaussian filter

[0126] Apply a first Gaussian filter to the filtered image Perform convolution to obtain a first convolution image;

[0127] Apply a second Gaussian filter to the filtered image Perform convolution to obtain a second convolution image;

[0128] Subtracting the first convolution image from the second convolution image to obtain a Gaussian difference image;

[0129] The pixel points that are 0 in the Gaussian difference image are taken to obtain a texture edge information image corresponding to the filtered image.

[0130] In one embodiment, performing connected domain processing on the texture edge information image to obtain a connected domain set includes:

[0131] Performing expansion processing on the texture edge information image to obtain an expanded texture image;

[0132] performing corrosion processing on the expanded texture image to obtain a corroded texture image;

[0133] All connected domains in the eroded texture image are found to obtain a connected domain set.

[0134] In one embodiment, scanning each connected domain in the connected domain set to obtain a corresponding edge contour region includes:

[0135] Filtering all connected domains in the connected domain set, removing connected domains whose areas are smaller than a preset area, and obtaining a standard connected domain set;

[0136] Contour scanning is performed on all connected domains in the standard connected domain set to obtain edge contour areas of opposite-sided edges corresponding to each connected domain in the standard connected domain set.

[0137] In one embodiment, the step of marking each edge contour area with a rectangular frame to obtain the corresponding defect area includes:

[0138] The corresponding minimum circumscribed slope distance shape is determined according to the polygonal edge of each edge contour area, and each minimum circumscribed slope distance shape is used as the defect area corresponding to each edge contour area.

[0139] Those skilled in the art will appreciate that all or part of the processes in the above-mentioned embodiments can be implemented by instructing the relevant hardware through a computer program. The computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to memory, storage, database or other media used in the embodiments provided in this application can include non-volatile and / or volatile memory. Non-volatile memory can include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM) or flash memory. Volatile memory can include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in various forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link (Synchlink) DRAM (SLDRAM), memory bus (Rambus) direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM).

[0140] The technical features of the above embodiments can be combined arbitrarily. To make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0141] The above description is merely a preferred embodiment of the present invention. The scope of protection of the present invention is not limited to the above embodiment. All technical solutions based on the concept of the present invention are within the scope of protection of the present invention. It should be noted that for those skilled in the art, various improvements and modifications that do not depart from the principles of the present invention should also be considered within the scope of protection of the present invention.

Claims

1. A texture-based defect detection method, characterized in that: include: Acquire image information of the grey fabric surface; performing guided filtering on the image information to obtain a filtered image; Extracting texture edge information of the filtered image to obtain a texture edge information image; Performing connected domain processing on the texture edge information image to obtain a connected domain set; Scanning each connected domain in the connected domain set to obtain a corresponding edge contour area; Mark each edge contour area with a rectangular frame to obtain the corresponding defect area; The performing guided filtering on the image information to obtain a filtered image includes: Obtaining a surface regular texture of the image information to obtain a guide picture; Performing guided filtering on the image information based on the guided image to obtain a filtered image; The step of extracting texture edge information of the filtered image to obtain a texture edge information image includes: Preset the variance σ1, σ2 and scaling ratios k1, k2 of the Gaussian convolution kernel to obtain the first Gaussian filter and the second Gaussian filter ; Apply a first Gaussian filter to the filtered image Perform convolution to obtain a first convolution image; Apply a second Gaussian filter to the filtered image Perform convolution to obtain a second convolution image; Subtracting the first convolution image from the second convolution image to obtain a Gaussian difference image; Taking the pixel points that are 0 in the Gaussian difference image to obtain the texture edge information image corresponding to the filtered image; The performing connected domain processing on the texture edge information image to obtain a connected domain set includes: Performing expansion processing on the texture edge information image to obtain an expanded texture image; performing corrosion processing on the expanded texture image to obtain a corroded texture image; Find all connected domains in the eroded texture image to obtain a connected domain set; Scanning each connected domain in the connected domain set to obtain a corresponding edge contour area includes: Filtering all connected domains in the connected domain set, removing connected domains whose areas are smaller than a preset area, and obtaining a standard connected domain set; Contour scanning is performed on all connected domains in the standard connected domain set to obtain edge contour areas of polygonal edges corresponding to each connected domain in the standard connected domain set.

2. The method according to claim 1, characterized in that The performing guided filtering on the image information based on the guided picture to obtain a filtered image includes: Input the image information, guide image, window radius r and regularization term Ɛ into the guided filtering algorithm; ， ， ， ； ， ； ， ； , ; .* + ; in, It is a time complexity of The mean filter is used; p is the image information, I is the guide image, and q is the output filtered image.

3. The method according to claim 1, characterized in that The rectangular frame marking of each edge contour area to obtain the corresponding defect area includes: The corresponding minimum circumscribed slope distance shape is determined according to the polygonal edge of each edge contour area, and each minimum circumscribed slope distance shape is used as the defect area corresponding to each edge contour area.

4. A texture-based defect detection system, characterized in that: include: An image acquisition module is used to acquire image information of the surface of the grey cloth; An image filtering module, configured to perform guided filtering on the image information to obtain a filtered image; A texture extraction module, configured to extract texture edge information of the filtered image to obtain a texture edge information image; A connected domain identification module is used to perform connected domain processing on the texture edge information image to obtain a connected domain set; An edge recognition module is used to scan each connected domain in the connected domain set to obtain a corresponding edge contour area; The defect marking module is used to mark each edge contour area with a rectangular frame to obtain the corresponding defect area; The texture-based defect detection system performs the steps of the method of any one of claims 1 to 3.

5. A computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the processor executes the computer program, the steps of the method according to any one of claims 1 to 3 are implemented.

6. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 3 are implemented.

Images