Post-correction methods, systems, devices, and media for small image based defect detection

Abstract

The application relates to a post-correction method, system, device and medium based on small picture-based defect detection, and the technical scheme points are as follows: including: acquiring defect positions in each cloth surface small picture; generating corresponding defect coordinates according to each defect position; generating a fitting straight line corresponding to all defect coordinates by using a least square method according to a defect type; establishing a defect frame according to the fitting straight line; and determining a defect length according to the defect frame; the application has the advantages that the defect detection result on the picture of the small picture can be used for post-correction, so that the accuracy of the length of the defects detected by the detection algorithm is improved.

Description

Technical Field

[0001] This invention relates to the field of fabric defect detection technology, and more specifically, to a post-correction method, system, device, and medium for defect detection based on small images. Background Technology

[0002] In the fabric production process, a weaving and inspection robot is generally used to weave the fabric while detecting defects on the fabric surface. Typically, a camera spanning across the fabric is used to capture images of small sections, resulting in several small images of the fabric surface. Defect detection is then performed on these small images to determine the location of defects in each small image.

[0003] In actual production, defects on all small images of the fabric surface need to be detected in order to obtain the final large image of the entire fabric surface. This detection method requires high recall and accuracy of the image recognition algorithm, and it is difficult to train a model with such high precision. Therefore, proposing a method for post-correction using the defect detection results on the small images has become a technical problem that urgently needs to be solved by those skilled in the art. Summary of the Invention

[0004] In view of the shortcomings of the existing technology, the purpose of this invention is to provide a post-correction method, system, device and medium for defect detection based on small images, which has the advantage of being able to use the defect detection results on the small image for post-correction, so as to improve the accuracy of the length of the defect detected by the detection algorithm.

[0005] The above-mentioned technical objective of the present invention is achieved through the following technical solution: a post-correction method for defect detection based on small images, comprising:

[0006] Obtain the location of defects in each fabric surface image;

[0007] Generate corresponding defect coordinates based on the location of each defect;

[0008] Based on the defect type, the least squares method is used to generate the fitted straight line corresponding to all defect coordinates;

[0009] A defect diagram frame is established based on the fitted straight line;

[0010] The length of the defect is determined based on the defect frame.

[0011] Optionally, obtaining the defect location in each fabric plot includes:

[0012] Defects are identified in each fabric panel image. The corresponding defects in each fabric panel image are marked with rectangles to obtain the location of the defects in each fabric panel image.

[0013] Optionally, the generating of the corresponding defect coordinates according to the respective defect positions comprises:

[0014] A plane coordinate system is established with the weft direction of the cloth as the X axis and the warp direction of the cloth as the Y axis.

[0015] The center point coordinates of the rectangular frame corresponding to each defect position are taken as the defect coordinates of the corresponding cloth face small graph.

[0016] Optionally, the generating of the fitting straight line corresponding to all defect coordinates by using the least square method according to the defect type comprises:

[0017] The defect types of the corresponding defects in each cloth face small graph are obtained.

[0018] For the warp direction defects, all defects with consistent defect types and weft direction coordinates floating within a preset range are taken to establish a warp direction defect coordinate point set; and a corresponding warp direction fitting straight line is generated by using the least square method according to the warp direction defect coordinate point set.

[0019] For the weft direction defects, all defects with consistent defect types and warp direction coordinates floating within a preset range are taken to establish a weft direction defect coordinate point set; and a corresponding weft direction fitting straight line is generated by using the least square method according to the weft direction defect coordinate point set.

[0020] Optionally, the establishing of the defect frame according to the fitting straight line comprises:

[0021] A midpoint of the intersection part of the fitting straight line and the cloth face small graph in the plane coordinate system is taken as the frame center point, an average value of the widths of the cloth face small graphs on both sides of the center point is taken as the frame width, a length of the intersection part of the fitting straight line and the cloth face small graph is taken as the frame length, and a corresponding defect frame is established.

[0022] Optionally, the determining of the defect length according to the defect frame comprises:

[0023] It is judged whether there are multiple defect positions in the defect frame, if yes, an area between two adjacent defect positions is regarded as a defect area, and if no, a rectangular frame corresponding to the defect position is regarded as a defect area.

[0024] The length of the defect area is calculated as the defect length.

[0025] Optionally, before the area between the two adjacent defect positions is regarded as the defect area, the method further comprises:

[0026] The confidence of each defect position in the defect frame is obtained.

[0027] If the confidence degrees of the positions of the two adjacent defects are both greater than the preset confidence threshold, the area between the positions of the two adjacent defects is regarded as a defect area.

[0028] A post-correction system for defect detection based on small pictures, comprising: a defect acquisition module, configured to acquire defect positions in each small picture of a cloth surface;

[0029] A coordinate generation module, configured to generate corresponding defect coordinates according to each defect position;

[0030] A linear fitting module, configured to generate a fitting straight line corresponding to all defect coordinates by using a least square method according to a defect type;

[0031] A frame drawing module, configured to establish a defect frame according to the fitting straight line;

[0032] A length calculation module, configured to determine a defect length according to the defect frame.

[0033] A computer device, comprising a memory and a processor, wherein the memory stores a computer program, and the processor implements the steps of the method when executing the computer program.

[0034] A computer readable storage medium, which stores a computer program, and the computer program implements the steps of the method when executed by a processor.

[0035] In summary, the present application has the following advantages: the post-correction is performed by using the defect detection result on the picture of the small picture of the cloth surface, for the small picture of the cloth surface on which no defect area is detected, the linear fitting is performed by using the least square method, the missed part is estimated by using the fitted straight line, and the missed part and the detected part are connected to splice the length of the defect completed on the gray cloth, thereby improving the length accuracy of the surface defect of the cloth, and further improving the defect accuracy in the gray cloth quality inspection report and the quality of the gray cloth defect quality inspection report. BRIEF DESCRIPTION OF DRAWINGS

[0036] Figure 1 The figure is a flowchart of the present application;

[0037] Figure 2 The figure is a structure block diagram of the present application when assembled;

[0038] Figure 3 The figure is an internal structure diagram of the computer device in the embodiment of the present application. DETAILED DESCRIPTION

[0039] In order to make the objects, features and advantages of the present application more clear, the specific embodiments of the present application are described in detail below with reference to the drawings. The drawings show several embodiments of the present application. However, the present application can be implemented in many different forms and is not limited to the embodiments described herein.

[0040] In the present application, unless specifically defined and limited otherwise, the terms "mounting", "connection", "connecting", "fixing" and the like should be interpreted broadly, for example, can be fixed connection, can also be detachable connection, or integrally connected; can be mechanical connection, can also be electrical connection; can be directly connected, can also be indirectly connected through an intermediate medium, can be the communication inside two elements. For those skilled in the art, the specific meaning of the above terms in the present application can be understood according to the specific circumstances. The terms "first", "second" are only for the purpose of description, and cannot be understood as indicating or implying relative importance or implicitly indicating the number of the indicated technical features. Therefore, the features limited by "first", "second" can be explicitly or implicitly included one or more features.

[0041] In the present application, unless specifically defined and limited otherwise, the first feature "on" or "under" the second feature can include that the first and second features are in direct contact, or that the first and second features are not in direct contact but are in contact through another feature between them. Moreover, the first feature "on", "above" and "over" the second feature includes that the first feature is directly above and obliquely above the second feature, or only indicates that the horizontal height of the first feature is higher than that of the second feature. The first feature "under", "below" and "under" the second feature includes that the first feature is directly below and obliquely below the second feature, or only indicates that the horizontal height of the first feature is less than that of the second feature. The terms "vertical", "horizontal", "left", "right", "up", "down" and similar expressions are only for the purpose of description, and are not indicative or suggestive of the devices or elements referred to must have a particular orientation, be constructed and operated in a particular orientation, and therefore cannot be understood as a limitation on the present application.

[0042] The present application is described in detail below with reference to the drawings and examples.

[0043] The present application provides a post-correction method for defect detection based on small maps, as shown in Figure 1 The present application provides a post-correction method for defect detection based on small maps, as shown in

[0044] Step 100, acquiring the defect position in each cloth small map;

[0045] Step 200, generating corresponding defect coordinates according to each defect position;

[0046] Step 300, generating a fitting straight line corresponding to all defect coordinates by using least square method according to the defect type;

[0047] Step 400, establishing a defect frame according to the fitting straight line;

[0048] Step 500, determining the defect length according to the defect frame.

[0049] In practical applications, all cloth surface small pictures are identified by an image recognition algorithm to obtain the defect positions in each cloth surface small picture. Then, each cloth surface small picture is spliced according to the shooting position of the cloth surface small picture to obtain a cloth surface large picture, and the defect coordinates corresponding to the defect positions in each cloth surface small picture after splicing are obtained. For example, the width of the cloth surface is 3 meters and the length is 450 meters. When shooting, the width of the cloth surface small picture obtained by the camera is generally greater than 0.5 meters and the length is generally 1 meter. Therefore, when splicing, generally 6 cloth surface small pictures in the width direction and 450 cloth surface small pictures in the length direction are spliced to form a cloth surface large picture. Then, since there are multiple defects of different types in the cloth surface large picture, the type of each defect is analyzed, and the least square method is used for linear fitting according to the type of each defect to obtain a fitting straight line corresponding to each defect. A defect frame is established according to the fitting straight line. Generally, the defect coordinates within the coordinate floating range in the preset range form a defect point set, and the least square method is used for linear fitting on the defect point set. The defect frame is established by the fitting straight line. Then, the length of the defect is determined according to the number of defects in the defect frame and the confidence of each defect.

[0050] Through the above scheme, the detected defects are linearly fitted, the missed part of the defects is estimated and detected by using the fitted straight line, and the missed part of the defects is connected with the detected defects. The length of the defect on the gray cloth is determined after splicing. Through this method, the length accuracy of the surface defects of the cloth can be improved, and the defect accuracy in the gray cloth quality inspection report can be improved, and the quality of the gray cloth defect quality inspection report can be improved.

[0051] Further, the defect positions in each cloth surface small picture are obtained, including:

[0052] The defects in each cloth surface small picture are identified, and the corresponding defects in each cloth surface small picture are labeled by a rectangular frame to obtain the defect positions in each cloth surface small picture.

[0053] In practical applications, each defect in the cloth surface small picture is labeled by a rectangular frame through an image recognition algorithm to obtain the defect positions in each cloth surface small picture. The edges of the rectangular frame are parallel to the edges of the cloth surface small picture.

[0054] Optionally, the corresponding defect coordinates are generated according to the defect positions, including:

[0055] A plane coordinate system is established with the weft direction of the cloth as the X axis and the warp direction of the cloth as the Y axis.

[0056] The center point coordinates of the rectangular frame corresponding to each defect position are taken as the defect coordinates of the corresponding cloth small picture.

[0057] In actual application, on the device of the edge weaving and inspection robot, the moving direction of the cloth is the warp direction, and the moving direction of the camera is the weft direction; a plane coordinate system is established with the weft direction of the cloth as the X axis and the warp direction of the cloth as the Y axis, and the center point of the rectangular frame corresponding to each defect position is converted into a coordinate point in the plane coordinate system, so that the defect coordinates of the corresponding defect in the corresponding cloth small picture are obtained.

[0058] Optionally, the generating of the fitting straight line corresponding to all defect coordinates by using the least square method according to the defect type comprises:

[0059] The defect type of the corresponding defect in each cloth small picture is obtained.

[0060] For the warp direction type defect, a warp direction defect coordinate point set is established for all defects with the same defect type and the weft direction coordinate floating range within a preset range; and a corresponding warp direction fitting straight line is generated by using the least square method according to the warp direction defect coordinate point set.

[0061] For the weft direction type defect, a weft direction defect coordinate point set is established for all defects with the same defect type and the warp direction coordinate floating range within a preset range; and a corresponding weft direction fitting straight line is generated by using the least square method according to the weft direction defect coordinate point set.

[0062] In actual application, different ways are adopted for linear fitting according to the defect type of the defect corresponding to each cloth small picture; for the warp direction type defect, a physical coordinate point set is generated for defects with the same defect type and the weft direction coordinate floating range within 5 cm, and a straight line equation y=kx+b is generated by using the least square method; and a corresponding warp direction fitting straight line is obtained; for the weft direction type defect, a physical coordinate point set is generated for defects with the same defect type and the warp direction coordinate floating range within 5 cm, and a straight line equation y=kx+b is generated by using the least square method; and a corresponding weft direction fitting straight line is obtained.

[0063] Further, the establishing of the defect frame according to the fitting straight line comprises:

[0064] The midpoint of the intersection part of the fitting straight line and the cloth small picture in the plane coordinate system is taken as the frame center point, the average width of the cloth small pictures on both sides of the center point is taken as the frame width, the length of the intersection part of the fitting straight line and the cloth small picture is taken as the frame length, and a corresponding defect frame is established.

[0065] In practical applications, the midpoint of the intersection part of the fitted straight line and all the cloth small pictures in the plane coordinate system is taken as the picture frame center point, and a defect picture frame is established based on the picture frame center point, so as to cover the defects on both sides of the fitted straight line.

[0066] Further, the determination of the defect length in the defect picture frame comprises:

[0067] If there are multiple defect positions in the defect picture frame, the area between the adjacent two defect positions is regarded as a defect area; if not, the rectangular frame corresponding to the defect position is regarded as a defect area.

[0068] The length of the defect area is calculated as the defect length.

[0069] In practical applications, there may be multiple defect positions in the defect picture frame, and the defect types of these defect positions are the same defect type. For example, a defect picture frame is generated by fitting a straight line in the weft direction, and there are two discontinuous rectangular frames corresponding to the defect coordinates in the defect picture frame. Therefore, the two discontinuous rectangular frames and the area between them are regarded as defect areas.

[0070] Further, before the area between the adjacent two defect positions is regarded as a defect area, the method further comprises:

[0071] The confidence of each defect position in the defect picture frame is obtained.

[0072] If the confidence of the adjacent two defect positions is greater than the preset confidence threshold, the area between the adjacent two defect positions is regarded as a defect area.

[0073] In practical applications, the defect identified by the image recognition algorithm has a corresponding confidence value. Therefore, when multiple defects are judged, the confidence of each defect position needs to be obtained. Only when the confidence of the adjacent two defect positions is greater than the preset confidence threshold, the area between the adjacent two defect positions can be regarded as a defect area.

[0074] In practical applications, the warp direction long defect mainly includes broken warp, double warp, loose warp, wrong warp and reed mark; the weft direction long defect mainly includes double weft, weft shrinkage, weft band and broken weft. Because the cloth area is relatively large, the cloth image needs to be shot in blocks. The method of the present application is to perform post-correction based on the defect detection result on the cloth small picture. For the part of the cloth small picture where no defect area is detected, the least square method is used for linear fitting, the fitted straight line is used for estimation of the missed part, and the missed part and the detected part are connected to splice out the length of the defect completed on the cloth.

[0075] As Figure 2 shown, the application also provides a post-correction system for defect detection based on small pictures, comprising:

[0076] a defect acquisition module 10, configured to acquire defect positions in each small picture of a cloth surface;

[0077] a coordinate generation module 20, configured to generate corresponding defect coordinates according to each defect position;

[0078] a linear fitting module 30, configured to generate a fitting straight line corresponding to all defect coordinates by using a least square method according to a defect type;

[0079] a frame drawing module 40, configured to establish a defect frame according to the fitting straight line;

[0080] a length calculation module 50, configured to determine a defect length according to the defect frame.

[0081] For specific limitations of the post-correction system for defect detection based on small pictures, refer to the limitations of the post-correction method for defect detection based on small pictures in the foregoing, which will not be repeated here. Each module of the post-correction system for defect detection based on small pictures can be realized by software, hardware, and combinations thereof, in whole or in part. Each module can be embedded in or independent of a processor in a computer device in hardware form, or stored in a memory in a computer device in software form, so as to be called and executed by a processor to perform the operations corresponding to each module.

[0082] In one embodiment, a computer device is provided, which can be a server, and an internal structure diagram thereof can be as Figure 3 shown. The computer device comprises a processor, a memory, a network interface, and a database connected through a system bus. The processor of the computer device is configured to provide computing and control capabilities. The memory of the computer device comprises a non-volatile storage medium and an 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 the computer program in the non-volatile storage medium to run. The computer program is executed by the processor to implement a post-correction method for defect detection based on small pictures.

[0083] Those skilled in the art can understand Figure 3 that the structure shown in the figure is only a block diagram of part of the structure related to the scheme of the present application, and does not constitute a limitation on the computer device to which the scheme of the present application is applied. A specific computer device can include more or fewer components than those shown in the figure, or combine certain components, or have a different arrangement of components.

[0084] In one embodiment, a computer device is provided, comprising a memory and a processor, the memory storing a computer program, and the processor implementing the following steps when executing the computer program: obtaining defect positions in each cloth surface small picture;

[0085] generating corresponding defect coordinates according to each defect position;

[0086] generating a fitting straight line corresponding to all defect coordinates by using a least square method according to a defect type;

[0087] establishing a defect frame according to the fitting straight line;

[0088] determining a defect length according to the defect frame.

[0089] In one embodiment, the obtaining of the defect positions in each cloth surface small picture comprises:

[0090] performing defect identification on each cloth surface small picture, marking corresponding defects in each cloth surface small picture by a rectangular frame, and obtaining the defect positions in each cloth surface small picture.

[0091] In one embodiment, the generating of the corresponding defect coordinates according to each defect position comprises:

[0092] establishing a plane coordinate system with a weft direction of the gray cloth as an X axis and a warp direction of the gray cloth as a Y axis;

[0093] taking a center point coordinate of a rectangular frame corresponding to each defect position as a defect coordinate of the corresponding cloth surface small picture.

[0094] In one embodiment, the generating of the fitting straight line corresponding to all defect coordinates by using the least square method according to the defect type comprises:

[0095] obtaining a defect type of corresponding defects in each cloth surface small picture;

[0096] for warp direction defects, establishing a warp direction defect coordinate point set of all defects with a consistent defect type and a weft direction coordinate floating range within a preset range, and generating a corresponding warp direction fitting straight line by using a least square method according to the warp direction defect coordinate point set;

[0097] for weft direction defects, establishing a weft direction defect coordinate point set of all defects with a consistent defect type and a warp direction coordinate floating range within a preset range, and generating a corresponding weft direction fitting straight line by using a least square method according to the weft direction defect coordinate point set.

[0098] In one embodiment, the establishing of the defect frame according to the fitting straight line comprises:

[0099] Taking the midpoint of the intersection part of the fitting straight line and the cloth surface small map in the plane coordinate system as the picture frame center point, and taking the average value of the width of the cloth surface small map on both sides of the center point as the picture frame width; taking the length of the intersection part of the fitting straight line and the cloth surface small map as the picture frame length, a corresponding defect picture frame is established.

[0100] In one embodiment, the determining the defect length according to the defect picture frame comprises:

[0101] Judging whether there are multiple defect positions in the defect picture frame, if yes, regarding the area between the adjacent two defect positions as a defect area; if no, regarding the rectangular frame corresponding to the defect position as a defect area.

[0102] Calculating the length of the defect area as the defect length.

[0103] In one embodiment, before the step of regarding the area between the adjacent two defect positions as a defect area, the method further comprises:

[0104] Obtaining the confidence degree of each defect position in the defect picture frame.

[0105] Judging whether the confidence degrees of the adjacent two defect positions are both greater than a preset confidence threshold, if yes, executing the step of regarding the area between the adjacent two defect positions as a defect area.

[0106] Those skilled in the art can understand that all or part of the processes in the above-mentioned embodiments can be completed by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer readable storage medium, and when executed, can include the processes of the above-mentioned embodiments. Any reference to memory, storage, database or other medium used in the embodiments provided by the present 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. As an illustration but not limitation, RAM is available in various forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (DDR SDRAM), 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), etc.

[0107] Any combination of the technical features in the above embodiments can be made, and for the sake of brevity, not all possible combinations of the technical features in the above embodiments are described, however, as long as the combination of the technical features does not exist contradictions, it should be considered as the scope of the present disclosure.

[0108] The above description is only the preferred embodiments of the present application, and the protection scope of the present application is not limited to the above embodiments. Any technical solutions falling within the concept of the present application shall be considered as the protection scope of the present application. It should be noted that, for ordinary skilled in the art, some improvements and refinements without departing from the principles of the present application, these improvements and refinements should also be considered as the protection scope of the present application.

Claims

1. A post-correction method for small image based defect detection, characterized in that, The method comprises the following steps: obtaining defect positions in each cloth surface small image; generating corresponding defect coordinates according to each defect position; generating a fitting straight line corresponding to all defect coordinates by using a least square method according to a defect type; establishing a defect frame according to the fitting straight line; determining a defect length according to the defect frame; the step of obtaining defect positions in each cloth surface small image comprises the following steps: performing defect identification on each cloth surface small image, marking corresponding defects in each cloth surface small image by using a rectangular frame, and obtaining defect positions in each cloth surface small image; the step of generating corresponding defect coordinates according to each defect position comprises the following steps: establishing a plane coordinate system with a weft direction of the gray cloth as an X axis and a warp direction of the gray cloth as a Y axis; taking a center point coordinate of a rectangular frame corresponding to each defect position as a defect coordinate of the corresponding cloth surface small image; the step of generating a fitting straight line corresponding to all defect coordinates by using a least square method according to a defect type comprises the following steps: obtaining a defect type of corresponding defects in each cloth surface small image; for a warp direction defect, establishing a warp direction defect coordinate point set of all defects with a consistent defect type and a weft direction coordinate floating range within a preset range; and generating a corresponding warp direction fitting straight line by using a least square method according to the warp direction defect coordinate point set; for a weft direction defect, establishing a weft direction defect coordinate point set of all defects with a consistent defect type and a warp direction coordinate floating range within a preset range; and generating a corresponding weft direction fitting straight line by using a least square method according to the weft direction defect coordinate point set; the step of establishing a defect frame according to the fitting straight line comprises the following steps: taking a midpoint of an intersection part of the fitting straight line and the cloth surface small image in the plane coordinate system as a frame center point, taking an average value of the width of the cloth surface small image on both sides of the center point as a frame width, taking a length of the intersection part of the fitting straight line and the cloth surface small image as a frame length, and establishing a corresponding defect frame; the step of determining a defect length according to the defect frame comprises the following steps: judging whether there are multiple defect positions in the defect frame, if yes, regarding a region between adjacent two defect positions as a defect region, and if no, regarding a rectangular frame corresponding to the defect position as a defect region; calculating a length of the defect region as a defect length.

2. The method of claim 1, wherein, Before the step of regarding the region between adjacent two defect positions as the defect region, the method further comprises the following steps: obtaining a confidence degree of each defect position in the defect frame; judging whether the confidence degrees of adjacent two defect positions are both greater than a preset confidence degree threshold, if yes, performing the step of regarding the region between adjacent two defect positions as the defect region.

3. A post-correction system for small map based defect detection, characterized in that, The system comprises: a defect obtaining module, configured to obtain defect positions in each cloth surface small image; a coordinate generating module, configured to generate corresponding defect coordinates according to each defect position; a linear fitting module, configured to generate a fitting straight line corresponding to all defect coordinates by using a least square method according to a defect type; a frame drawing module, configured to establish a defect frame according to the fitting straight line; a length calculating module, configured to determine a defect length according to the defect frame; a post-correction system based on small image-based defect detection is used to perform the steps of the method in claim 1 or 2.

4. A computer device comprising a memory and a processor, the memory storing a computer program, characterized in that, The computer program is executed by the processor to implement the steps of the method of claim 1 or 2.

5. A computer-readable storage medium having stored thereon a computer program, characterized in that, The computer program is executed by the processor to implement the steps of the method of claim 1 or 2.

Images