Display screen repeated defect detection method and device and storage medium

Abstract

The invention discloses a method and device for detecting repeated defects of a display screen and a storage medium. The method and device are used for improving the detection efficiency of the defects of the display screen. Performing consistency preprocessing on the acquired display screen image; performing point-line block mura defect extraction processing on the display screen image, and performing defect region segmentation extraction processing to generate a point-line block defect region; performing blob analysis on the point-line block defect area to generate defect feature information; according to the aspect ratio AR value of the defect feature information and the defect position information, carrying out region combination on the point-line block defect region; for a plurality of display screen images of the same ID, calculating the adjacent matching degree of geometric positions according to the adjacent matching degree and the Euclidean distance of the defect areas of the two point-line blocks on every two display screen images; and performing display screen over-detection analysis processing according to the geometric position adjacency matching degree.

Description

Technical Field

[0001] The embodiments of the present application relate to the field of display screen detection, and in particular to a method, device and storage medium for detecting repeated defects in a display screen. Background Art

[0002] With the rapid development of industrial and Internet technologies, the industry's requirements for industrial product quality are becoming more and more refined and standardized. As far as the display panel industry is concerned, the quality of products directly affects its product competitiveness. Among them, the accuracy of product shape and size specifications is the most basic process indicator of product quality. As one of the display components of the device, the display screen is used in various high-end devices, such as mobile phones, TVs, tablets, etc. As people's requirements for screen display continue to increase, the display screen has gradually become a technologically sophisticated product.

[0003] Defect detection of display screens is a necessary process for adjustment. Display screen defects can be divided into display defects and screen body defects according to whether the display screen needs to be lit for detection. Display defects refer to the discordant areas that appear in the display screen during the display screen display process. In the prior art, display defects usually require inputting a specific picture on the display screen to be tested, and then collecting the picture through the camera, and then performing image detection to analyze it. Screen body defects refer to areas where damage occurs on the surface or inside of the display screen itself (such as broken corners, foreign matter, etc.). Such defects do not require inputting images. They can be analyzed by collecting the picture through the camera and then performing image detection through the automatic optical detection module. Some of these defects appear repeatedly in different grayscale images or even in different product photos, resulting in a large number of over-inspection phenomena. In order to reduce the over-inspection phenomenon, a defect area position matching algorithm is usually used for the defect area of ​​the display screen image to determine the same defect area on the display screen image of the same ID.

[0004] However, there are some linear scratch defects in the existing display screen defects. When shooting from a side angle, the scratches will be discontinuous, that is, there are breakpoint defects that belong to the same defect on the display screen image, which increases the chances of the display screen passing inspection. Secondly, in order to protect the display screen from being scratched by the robotic arm during displacement, a layer of protective film is usually attached to the surface of the display screen, which causes scratch defects to appear on the protective film. Although the protective film is set as thin as possible, it still has a certain thickness, and the scratches produced are deeper than the display screen, resulting in breakpoint defects that are more likely to occur. In addition, there may also be interlayer dirt defects between the protective film and the display screen, which increases the possibility of passing inspection and reduces the efficiency of detecting display screen defects. Summary of the invention

[0005] The present application discloses a method, device and storage medium for detecting repeated defects of a display screen, which are used to improve the detection efficiency of defects of the display screen.

[0006] The first aspect of the present application discloses a method for detecting repeated defects of a display screen, comprising: Perform consistency preprocessing on the collected display screen images; Perform point-line-block mura defect extraction on the display screen image, and perform defect area segmentation and extraction to generate point-line-block defect areas; Perform blob analysis on the point, line, and block defect areas to generate defect feature information; Merge the defect areas of the point and line blocks according to the aspect ratio AR value of the defect feature information and the defect position information; For several display screen images with the same ID, the geometric position adjacency matching degree is calculated based on the adjacency matching degree and Euclidean distance of the defect areas of two point-line blocks on each two display screen images; The display screen is inspected and analyzed based on the geometric position adjacency matching degree.

[0007] Optionally, after performing display screen over-inspection analysis processing according to the geometric position adjacency matching degree, the method further includes: For several display screen images with different IDs, calculate the geometric position relationship information and defect feature relationship information of two point-line block defect areas on every two display screen images; Batch anomaly detection of display screens is performed based on geometric position relationship information and defect feature relationship information.

[0008] Optionally, the geometric position relationship information is Euclidean distance information; For several display screen images with different IDs, the geometric position relationship information and defect feature relationship information of the two point-line block defect areas on every two display screen images are calculated, including: For several display screen images with different IDs, calculate the Euclidean distance information of two point-line block defect areas on every two display screen images; Defect feature relationship information is calculated based on the roundness calculation value, contrast calculation value and texture feature calculation value of the two dot-line block defect areas on every two display screen images.

[0009] Optionally, the point-line-block defect regions are merged according to the aspect ratio AR value of the defect feature information and the defect position information, including: Determine the linear defect area and the non-linear defect area in the point-line block defect area according to the aspect ratio AR value in the defect feature information; When the linear defect area is within the minimum circumscribed rectangular area of ​​the non-linear defect area, the Euclidean distance between the linear defect area and the non-linear defect area is calculated; When the linear defect region is not located within the minimum circumscribed rectangular region of the non-linear defect region, determining the orientation information of the linear defect region and the non-linear defect region; Calculate the Euclidean distance between the positions of the nearest corner points of the minimum circumscribed rectangle of the linear defect area in the non-linear defect area according to the orientation information; Merge analysis was performed using Euclidean distance; According to the analysis results, the point, line and block defect areas are merged and split to obtain the display screen image with connected domain labels, and the Blob information and feature information of the corresponding positions after the merger are generated.

[0010] Optionally, the collected display screen image is preprocessed for consistency, including: Perform color channel merging or splitting processing on the collected display screen images; Perform geometric correction on the display screen image; Pre-process and filter the display screen image to remove noise and background interference; The preprocessed image is enhanced, and the enhancement processing includes shadow correction and brightness enhancement.

[0011] Optionally, for several display screen images with the same ID, the geometric position adjacency matching degree is calculated based on the adjacency matching degree of the defective areas of two point-line blocks on each two display screen images and the Euclidean distance, including: For several display screen images with the same ID, determine the independent area area, the intersecting area area and the minimum circumscribed rectangular area of ​​the two defect areas on every two display screen images; Calculate the adjacency matching degree of the defect area according to the area of ​​the independent area, the area of ​​the intersecting area and the area of ​​the minimum circumscribed rectangle; Calculate the Euclidean distance of the defective area according to the coordinate information of the two defective areas on each two display screen images; The geometric position adjacency matching degree is calculated based on the defect area adjacency matching degree and the Euclidean distance of the defect area.

[0012] The second aspect of the present application discloses a device for detecting repeated defects of a display screen, comprising: A processing unit, used for performing consistency preprocessing on the collected display screen images; The first generating unit is used to perform point line block mura defect extraction processing on the display screen image, and perform defect area segmentation extraction processing to generate point line block defect areas; The second generating unit is used to perform blob analysis on the point line block defect area to generate defect feature information; The defect merging unit merges the defect areas of the point, line and block according to the aspect ratio AR value of the defect feature information and the defect position information; The first calculation unit is used to calculate the geometric position adjacency matching degree according to the adjacency matching degree and the Euclidean distance of the defect areas of two point-line blocks on each two display screen images for a plurality of display screen images with the same ID; The analysis unit is used to perform display screen over-inspection analysis and processing according to the geometric position adjacency matching degree.

[0013] Optionally, after the calculation unit, the device further includes: The second calculation unit is used to calculate the geometric position relationship information and defect feature relationship information of two point-line block defect areas on every two display screen images for a plurality of display screen images with different IDs; The detection unit is used to perform batch abnormality detection of display screens based on geometric position relationship information and defect feature relationship information.

[0014] Optionally, the geometric position relationship information is Euclidean distance information; The second computing unit comprises: For several display screen images with different IDs, calculate the Euclidean distance information of two point-line block defect areas on every two display screen images; Defect feature relationship information is calculated based on the roundness calculation value, contrast calculation value and texture feature calculation value of the two dot-line block defect areas on every two display screen images.

[0015] Optional, defect merging unit, including: Determine the linear defect area and the non-linear defect area in the point-line block defect area according to the aspect ratio AR value in the defect feature information; When the linear defect area is within the minimum circumscribed rectangular area of ​​the non-linear defect area, the Euclidean distance between the linear defect area and the non-linear defect area is calculated; When the linear defect region is not located within the minimum circumscribed rectangular region of the non-linear defect region, determining the orientation information of the linear defect region and the non-linear defect region; Calculate the Euclidean distance between the positions of the nearest corner points of the minimum circumscribed rectangle of the linear defect area in the non-linear defect area according to the orientation information; Merge analysis was performed using Euclidean distance; According to the analysis results, the point, line and block defect areas are merged and split to obtain the display screen image with connected domain labels, and the Blob information and feature information of the corresponding positions after the merger are generated.

[0016] Optionally, a processing unit comprising: Perform color channel merging or splitting processing on the collected display screen images; Perform geometric correction on the display screen image; Pre-process and filter the display screen image to remove noise and background interference; The preprocessed image is enhanced, and the enhancement processing includes shadow correction and brightness enhancement.

[0017] Optionally, the first computing unit includes: For several display screen images with the same ID, determine the independent area area, the intersecting area area and the minimum circumscribed rectangular area of ​​the two defect areas on every two display screen images; Calculate the adjacency matching degree of the defect area according to the area of ​​the independent area, the area of ​​the intersecting area and the area of ​​the minimum circumscribed rectangle; Calculate the Euclidean distance of the defective area according to the coordinate information of the two defective areas on each two display screen images; The geometric position adjacency matching degree is calculated based on the defect area adjacency matching degree and the Euclidean distance of the defect area.

[0018] A third aspect of the present application provides a device for detecting repeated defects of a display screen, comprising: Processor, memory, input-output unit, and bus; The processor is connected to the memory, the input and output unit, and the bus; The memory stores a program, and the processor calls the program to execute the first aspect and any optional method of the first aspect.

[0019] A fourth aspect of the present application provides a computer-readable storage medium, on which a program is stored. When the program is executed on a computer, the program executes the first aspect and any optional method of the first aspect.

[0020] It can be seen from the above technical solutions that the embodiments of the present application have the following advantages: In this application, the collected display screen images are firstly preprocessed for consistency, so that each image can avoid large display deviations and relative position deviations as much as possible. Each display screen image is subjected to point line block mura defect extraction processing, and defect area segmentation and extraction processing is performed to generate point line block defect areas, that is, defects of different forms are extracted and located. Next, the point line block defect area is subjected to blob analysis to generate defect feature information of each point line block defect area. The aspect ratio AR value is determined according to the defect feature information, and the point line block defect areas that meet the conditions are merged according to the aspect ratio AR value and the defect position information to reduce the breakpoint defects on the display screen or protective film on the same display screen image. For several display screen images with the same ID, the geometric position adjacency matching degree is calculated based on the adjacency matching degree and Euclidean distance of the two point line block defect areas on each two display screen images. The display screen is then inspected and analyzed based on the geometric position adjacency matching degree. Through the two parameters of adjacency matching and Euclidean distance, the geometric position adjacency matching of the defect area is analyzed in terms of position relationship and morphological relationship, making the image repeated defect detection more accurate and improving the detection efficiency of display screen defects. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative labor.

[0022] Figure 1 A schematic diagram of an embodiment of a method for repeated defect detection of a display screen of the present application; Figure 2 A schematic diagram of an embodiment of a method for batch abnormality detection of display screens of the present application; Figure 3 A schematic diagram of an embodiment of a method for calculating geometric position relationship information and defect feature relationship information in the present application; Figure 4 A schematic diagram of an embodiment of the method for merging regions of the present application; Figure 5 A schematic diagram of an embodiment of a method for consistency preprocessing of the present application; Figure 6 A schematic diagram of an embodiment of a method for calculating geometric position adjacency matching degree in the present application; Figure 7 A schematic diagram of an embodiment of a device for repeated defect detection of a display screen according to the present application; Figure 8A schematic diagram of another embodiment of the device for repeated defect detection of a display screen according to the present application. DETAILED DESCRIPTION

[0023] In the following description, specific details such as specific system structures, technologies, etc. are provided for the purpose of illustration rather than limitation, so as to provide a thorough understanding of the embodiments of the present application. However, it should be clear to those skilled in the art that the present application may also be implemented in other embodiments without these specific details. In other cases, detailed descriptions of well-known systems, devices, circuits, and methods are omitted to prevent unnecessary details from obstructing the description of the present application.

[0024] It should be understood that when used in the present specification and the appended claims, the term "comprising" indicates the presence of described features, wholes, steps, operations, elements and / or components, but does not exclude the presence or addition of one or more other features, wholes, steps, operations, elements, components and / or combinations thereof.

[0025] It should also be understood that the term “and / or” used in the specification and appended claims refers to any and all possible combinations of one or more of the associated listed items, and includes these combinations.

[0026] As used in the specification and appended claims of this application, the term "if" can be interpreted as "when" or "uponce" or "in response to determining" or "in response to detecting", depending on the context. Similarly, the phrase "if it is determined" or "if [described condition or event] is detected" can be interpreted as meaning "uponce it is determined" or "in response to determining" or "uponce [described condition or event] is detected" or "in response to detecting [described condition or event]", depending on the context.

[0027] In addition, in the description of the present application specification and the appended claims, the terms "first", "second", "third", etc. are only used to distinguish the descriptions and cannot be understood as indicating or implying relative importance.

[0028] References to "one embodiment" or "some embodiments" etc. described in the specification of this application mean that one or more embodiments of the present application include specific features, structures or characteristics described in conjunction with the embodiment. Therefore, the statements "in one embodiment", "in some embodiments", "in some other embodiments", "in some other embodiments", etc. that appear in different places in this specification do not necessarily refer to the same embodiment, but mean "one or more but not all embodiments", unless otherwise specifically emphasized in other ways. The terms "including", "comprising", "having" and their variations all mean "including but not limited to", unless otherwise specifically emphasized in other ways.

[0029] In the prior art, there are some linear scratch defects in display screen defects. When shooting from a side angle, the scratches will be discontinuous, that is, there are breakpoint defects that belong to the same defect on the display screen image, which increases the chances of the display screen passing inspection. Secondly, in order to protect the display screen from being scratched by the robotic arm during displacement, a layer of protective film is usually attached to the surface of the display screen, which causes scratch defects to appear on the protective film. Although the protective film is set as thin as possible, it still has a certain thickness, and the scratches produced are deeper than the display screen, resulting in breakpoint defects that are more likely to occur. In addition, there may also be interlayer dirt defects between the protective film and the display screen, which increases the possibility of passing inspection and reduces the efficiency of detecting display screen defects.

[0030] Based on this, the present application discloses a method, device and storage medium for detecting repeated defects in a display screen, which are used to improve the detection efficiency of defects in the display screen.

[0031] The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.

[0032] The method of the present application can be applied to a server, a device, a terminal or other devices with logic processing capabilities, and the present application does not limit this. For the convenience of description, the following description is made by taking the execution subject as an example of a terminal.

[0033] See also Figure 1 The present application provides an embodiment of a method for detecting repeated defects of a display screen, comprising: 101. Perform consistency preprocessing on the collected display screen image; The terminal first obtains a batch of display screens of the same type, and captures multiple images of different grayscales for each display screen with each ID. The images captured at different grayscales of the same display screen and the images captured at different grayscales of different display screens are pre-processed for consistency, so that images that cannot be compared with each other will not appear, and the position of the display screen can be calibrated, so that the position of the defective area is relatively accurate.

[0034] 102. Performing point-line-block mura defect extraction processing on the display screen image, and performing defect area segmentation extraction processing to generate point-line-block defect areas; The terminal extracts mura defects in points, lines, and blocks from the display screen image after consistency preprocessing. By extracting defects from small to large, it can accurately divide the defects, and use the point-line convolution filter window to segment and extract the defect area to generate point-line-block defect areas.

[0035] 103. Perform blob analysis on the point, line, and block defect areas to generate defect feature information; First, the terminal performs blob analysis on the extracted point, line, and block defect areas to extract defect feature information, including the coordinates of the center of gravity, upper left and lower right coordinates, corner coordinates, area, minimum circumscribed rectangle boundary coordinates, aspect ratio, foreground (background) average concentration, contrast, and shape and texture features of the corresponding defect area.

[0036] 104. Merge the defect areas of the point and line blocks according to the aspect ratio AR value of the defect feature information and the defect position information; In this embodiment, the terminal merges the dot-line-block defect areas according to the aspect ratio AR value of the defect feature information and the defect position information to reduce the breakpoint defects on the display screen or the protective film on the same display screen image.

[0037] 105. For a plurality of display screen images with the same ID, the geometric position adjacency matching degree is calculated according to the adjacency matching degree of the defective areas of two point-line blocks on every two display screen images and the Euclidean distance; For several display screen images with the same ID, the terminal calculates the geometric position adjacency matching degree based on the adjacency matching degree and Euclidean distance of the defective areas of two dot-line blocks on each two display screen images. The geometric position adjacency matching degree of the defective area can be analyzed in terms of positional relationship and morphological relationship through the two parameters of adjacency matching degree and Euclidean distance, making the image repeated defect detection more accurate and improving the detection efficiency of display screen defects.

[0038] 106. Perform display screen inspection analysis and processing based on geometric position adjacency matching.

[0039] Finally, the terminal determines multiple areas of the same defect in the display screen image with the same ID based on the geometric position adjacency matching degree, performs display screen inspection and analysis, and reduces the occurrence of repeated confirmation of defective areas.

[0040] When repeated defects are detected with the same defect ID, according to the set repeated defect threshold parameters, when the field application encounters quality abnormal defects, the defect at the current location can be set as one defect, without repeated alarm card control, to prevent repeated alarms and increase redundant information. The field user only needs to observe the final defect information; In this embodiment, the collected display screen images are firstly subjected to consistency preprocessing so that each image can avoid large display deviations and relative position deviations as much as possible. Each display screen image is subjected to point line block mura defect extraction processing, and defect area segmentation and extraction processing is performed to generate point line block defect areas, that is, defects of different forms are extracted and located. Next, the point line block defect area is subjected to blob analysis to generate defect feature information of each point line block defect area. The aspect ratio AR value is determined according to the defect feature information, and the point line block defect areas that meet the conditions are merged according to the aspect ratio AR value and the defect position information to reduce the breakpoint defects on the display screen or protective film on the same display screen image. For several display screen images with the same ID, the geometric position adjacency matching degree is calculated based on the adjacency matching degree and Euclidean distance of the two point line block defect areas on each two display screen images. Then, the display screen is inspected and analyzed according to the geometric position adjacency matching degree. Through the two parameters of adjacency matching and Euclidean distance, the geometric position adjacency matching of the defect area is analyzed in terms of position relationship and morphological relationship, making the image repeated defect detection more accurate and improving the detection efficiency of display screen defects.

[0041] See also Figure 2 The present application provides an embodiment of a method for batch abnormality detection of display screens, comprising: 201. For a plurality of display screen images with different IDs, calculate geometric position relationship information and defect feature relationship information of two point-line block defect areas on every two display screen images; 202. Perform batch anomaly detection on display screens based on geometric position relationship information and defect feature relationship information.

[0042] In this embodiment, the terminal first calculates the geometric position relationship information and defect feature relationship information of two point-line block defect areas on each two display screen images for several display screen images with different IDs. The collective position relationship refers to the relative position of the two defect areas, and the defect feature relationship information refers to the similarity and correlation between the morphological feature data on the defect areas.

[0043] The terminal performs batch abnormality detection of display screens based on the geometric position relationship information and defect feature relationship information. Specifically, during the production and inspection process of display screens with different IDs, the industrial computer of each machine on site detects in real time whether there are repeated defects. If there are detections and the number of repetitions exceeds the number of experience set by the user, the process alarm prompts at this time to remind in time to prevent batch abnormalities of the current machine; in addition, the repeated defect data information generated each time is stored in a given format with parameters, and is stored as a database file with less information, small file space, fast upload and download parsing speed, and conciseness. Therefore, this type of data can be uploaded to the workshop shared directory for all machines to share. CIM and other third-party software can achieve timely data sharing and timely early warning decisions by real-time monitoring of repeated defect generation data.

[0044] After the detection is completed, defect integration is carried out. The steps are as follows: 1) During the first run, a repeated defect data set is generated and stored in the database. In addition to the coordinate information of the current defect and the defect information features (position coordinates, area, shape, texture, contrast value, etc.) of the previous step, the data set also contains the number of consecutive repetitions m and the number of repetitions mn in the given m data; 2) When the subsequent product is running, if the geometric positions and features of the two detected defects can match, they are regarded as repeated defect positions, and the repeated defect information at this time is recorded, and the number of repetitions m is increased by 1, and the number of nm is also increased by 1. If no repeated defects are found, the current data set m is set to 0, and the number of nm remains unchanged; 3) According to the software settings, it can be divided into continuous counting method and interval counting method. a. If the continuous counting method is set, each time it is determined that there is a repeated defect, the current defect data information will be stored in the database according to the specified format, and the number of repetitions m will be increased by 1. If it is not a repeated defect, m is set to 0. This is the same for each product inspection process; b. If the interval counting method is set, first set the maximum number of storage entries, such as 100 (when the production data is less than 100, the extra entries are set to be empty or the default value). Each entry contains a piece of defect information detected by the product: When the data is less than 100, the defect information of each input data is stored in the entry in order. Each time it is judged to be a repeated defect, the corresponding nm flag is increased by 1, otherwise it remains unchanged.

[0045] When the number of production data exceeds 100, the latest entry in the data set replaces the earliest entry.

[0046] See also Figure 3 The present application provides an embodiment of a method for calculating geometric position relationship information and defect feature relationship information, including: 301. For a plurality of display screen images with different IDs, calculate the Euclidean distance information of two point-line block defect areas on every two display screen images; 302. Calculate defect feature relationship information according to the roundness calculation value, contrast calculation value and texture feature calculation value of the two point-line block defect areas on every two display screen images.

[0047] The terminal calculates the Euclidean distance information of the two dot-line block defect areas on each of the two display screen images for several display screen images with different IDs. The defect feature relationship information is calculated based on the roundness calculation value, contrast calculation value and texture feature calculation value of the two dot-line block defect areas on each of the two display screen images. Specifically, for different types of product IDs, the terminal will first record the defect information of the two running display screens, including the defect type (bright spot or dark spot or other), area, center of gravity coordinates, coordinates of the upper left and lower right corners of the minimum circumscribed rectangle, texture features, contrast, shape feature information, etc.

[0048] When the subsequent display screen is running, the current defect data information is compared with the data of the previous display screen to consider whether it is a repeated defect from two dimensions, one is the geometric position relationship, and the other is the feature value matching degree: if the geometric position and features of the two detected defects can match, it is regarded as a repeated defect position, and the repeated defect information at this time is recorded. In terms of geometric relationship, the distance between the coordinates of the respective defect centers can be used, and the feature values ​​are matched from the contrast, texture, and shape features: The geometric relationship characteristics are as follows:

[0049] is the Euclidean distance between the defect p region and the defect q region, is the abscissa of the center coordinate of the defect p region, is the ordinate of the center coordinate of the defect p region, is the abscissa of the center coordinate of the defect q region, It is the ordinate of the center coordinate of the defect q area.

[0050] The simplified expression of the characteristic value (defect characteristic relationship information) is:

[0051] in represents the calculated value of roundness, is the contrast calculation value, The calculated value is the texture feature. The calculated value of roundness refers to the roundness morphological feature of the defect area. Even if there is a protective film, the corresponding calculated value of roundness is still generated based on the image information.

[0052] First, obtain the P defect area , and , to obtain the Q defect area , and , the specific formula is as follows: First, calculate the weight of the roundness calculation value, the contrast calculation value, and the texture feature calculation value. The calculation method is as follows:

[0053]

[0054]

[0055]

[0056] In the above formula, the characteristic quantities of three defect areas, namely the roundness calculation value, contrast calculation value and texture feature calculation value, are used, which are integrated to form the weight parameters between the defect areas, and then the defect feature relationship information is calculated, which can better compare the morphological correlation of the defect areas.

[0057] See also Figure 4 The present application provides an embodiment of a method for merging regions, including: 401. Determine a line defect region and a non-line defect region in a point-line block defect region according to an aspect ratio AR value in defect feature information; 402. When the linear defect region is within the minimum circumscribed rectangular region of the non-linear defect region, calculate the Euclidean distance between the linear defect region and the non-linear defect region; 403. When the linear defect region is not located within the minimum circumscribed rectangular region of the non-linear defect region, determine the orientation information of the linear defect region and the non-linear defect region; 404. Calculate the Euclidean distance between the positions of the nearest corner points of the minimum circumscribed rectangle of the linear defect region in the non-linear defect region according to the orientation information; 405. Merge analysis through Euclidean distance; 406. Merge and split the point, line, and block defect areas according to the analysis results, obtain a display screen image with a connected domain label, and generate Blob information and feature information of the corresponding position after the merger.

[0058] In this embodiment, the terminal determines the line - type defect area and the non - line - type defect area in the dot - line block defect area according to the aspect ratio AR value in the defect feature information. When the line - type defect area is within the minimum circumscribed rectangle area of the non - line - type defect area, calculate the Euclidean distance between the line - type defect area and the non - line - type defect area. When the line - type defect area is not within the minimum circumscribed rectangle area of the non - line - type defect area, determine the orientation information of the line - type defect area and the non - line - type defect area. Calculate the Euclidean distance of the line - type defect area at the position of the similar corner point of the minimum circumscribed rectangle of the non - line - type defect area according to the orientation information. The terminal performs merging analysis through the Euclidean distance. The terminal merges and splits the dot - line block defect area according to the analysis result, obtains the display screen image with connected - domain labels, and generates the Blob information and feature information at the corresponding positions after merging.

[0059] Specifically, the terminal merges the defect areas of the segmented regions and integrates the regions according to the blob region information, aiming to reduce the false detection of the blob regions after the segmentation of small defect regions. Ideally, a single defect target should be connected after segmentation. However, due to factors such as the setting of the defect segmentation algorithm threshold and uneven display screen brightness, the same defect area may be segmented into multiple small sub - defects, which will cause the feature quantity of the defect to be distorted during subsequent repeated foreign object judgment, resulting in misjudgment of repeated defects. Therefore, this embodiment proposes an algorithm for overall analysis and merging of the defect extraction regions to ensure the accuracy of subsequent repeated foreign object extraction: When the terminal performs blob analysis on the extracted dot - line block defect area and extracts the defect feature information, it obtains the centroid coordinates, upper - left and lower - right coordinates, corner coordinates, area, minimum circumscribed rectangle boundary coordinates, aspect ratio, front (background) average concentration, contrast, and shape texture features of the corresponding defect area.

[0060] Next, the terminal calculates the regions geometrically close to the current defect area, and the steps are as follows: a. Suppose k defects are detected, and the current defect serial number is i (0 < i <= k). Determine whether the current defect area p is a line - type defect. Consider the defect with an aspect ratio AR value greater than the given parameter as a line - type abnormality, otherwise it is a non - line - type defect; b. The merging strategy for line - type defects and non - line defects is as follows. For the line - type defect p defect area, first determine whether the p defect area belongs to the minimum circumscribed rectangle area of other non - line - type defect i. If it is within the area, only calculate the Euclidean distance from the p defect area to the i defect area. If it is not within the minimum circumscribed rectangle area, continue the following determination; c. Determine which position (upper - left, lower - left, upper - right, lower - right) the p defect area is in the i defect area, and then calculate the Euclidean distance between the p defect area and the position of the similar corner point of the minimum circumscribed rectangle of the i defect area; d. Then the terminal classifies the defective areas that can be merged into one category according to the calculated Euclidean distance, and then sets a label number for each category of areas; Based on the previous step, the terminal further merges and splits the regions through texture, shape, contrast and other feature quantities. Finally, the label map with connected domain labels is obtained, and the Blob information and feature information of the corresponding position after merging are calculated. Then the defect feature information of the segmented region is obtained, including feature information such as area, coordinate position, shape feature (horizontal and vertical length), and contrast.

[0061] See also Figure 5 The present application provides an embodiment of a consistency preprocessing method, comprising: 501. Perform color channel merging or splitting processing on the collected display screen image; 502. Perform geometric correction processing on the display screen image; 503. Preprocess and filter the display screen image to remove noise and background interference; 504. Perform enhancement processing on the preprocessed image, where the enhancement processing includes shadow correction and brightness enhancement.

[0062] In this embodiment, the terminal performs color channel merging or splitting processing on the collected display screen image, and then performs geometric correction processing on the display screen image. The terminal performs preprocessing and filtering on the display screen image to remove noise and background interference, and finally enhances the preprocessed image, and the enhancement processing includes shadow correction and brightness enhancement. Specifically, the terminal performs color channel merging or splitting processing on the display screen image to generate a grayscale image or a single-channel image. The terminal then performs geometric correction on the above image, in order to be able to use the corner point relative to the screen resolution as the origin when calculating the coordinates later. For different products, due to the change in the position of the incoming material placement, it is very important to calculate the relative coordinates of the same standard. Next, the terminal performs preprocessing and filtering on the geometrically corrected display screen image to remove noise, background and other interference. Finally, the terminal further enhances the preprocessed display screen image, including shadow correction, brightness enhancement, etc.

[0063] See also Figure 6 The present application provides an embodiment of a method for calculating a geometric position adjacency matching degree, comprising: 601. For a plurality of display screen images with the same ID, determine the independent area area, the intersecting area area and the minimum circumscribed rectangular area of ​​two defect areas on every two display screen images; 602. Calculate the adjacency matching degree of the defect area according to the area of ​​the independent area, the area of ​​the intersecting area and the area of ​​the minimum circumscribed rectangle; 603. Calculate the Euclidean distance of the defective areas according to the coordinate information of the two defective areas on every two display screen images; 604. Calculate the geometric position adjacency matching degree according to the defect area adjacency matching degree and the Euclidean distance of the defect area.

[0064] In this embodiment, the terminal first determines the independent area areas of the two defective areas on every two display screen images for a plurality of display screen images with the same ID. and , the area of ​​the intersection The minimum circumscribed rectangular area of ​​the defect area composed of Next, the terminal calculates the adjacency matching degree of the defective area according to the area of ​​the independent area, the area of ​​the intersecting area and the area of ​​the minimum circumscribed rectangle. The Euclidean distance of the defective area is calculated according to the coordinate information of the two defective areas on each two display screen images. The geometric position adjacency matching degree is calculated according to the adjacency matching degree of the defective area and the Euclidean distance of the defective area.

[0065] Specifically, for the same product ID, the position of the defect area generally remains roughly unchanged, but due to changes in the detection screen hardware, exposure, gain, etc. and differences in software detection parameters, there will be slight differences in the detection area. After region extraction and merging, the degree of similarity and matching of the defect area detected by the before and after comparison images is calculated.

[0066] After the two display screen images have obtained the merged defect area, the defect area adjacency matching degree ATR of the two areas r and t between the two display screen images is calculated:

[0067] Then, the Euclidean distance of the defective area is calculated based on the coordinate information of the two defective areas on each two display screen images. The formula is as follows:

[0068] is the Euclidean distance between defect r area and defect t area, is the abscissa of the center coordinate of the defect r region, is the ordinate of the center coordinate of the defect r area, is the abscissa of the center coordinate of the defect t area, It is the ordinate of the center coordinate of the defect t area.

[0069] The geometric position adjacency matching degree is expressed as IR, and the simplified formula is as follows:

[0070]

[0071]

[0072] Among them and represent the areas of regions r and t respectively, represents the intersection of two regions. Indicates the adjacency matching degree of two defect areas, generally using the Dice matching coefficient; represents the Euclidean distance between the center coordinates of r and t, The minimum enclosing rectangle area of ​​the area composed of r and t is represented by the above expression. When r and t satisfy as well as When the threshold parameters (a, b, c and d are preset threshold parameters) are met, it is considered as a candidate area for repeated defects.

[0073] Calculate whether the two defect feature information r and t can match. If they are close and the calculation result of the previous step is within the given parameter range, it is finally determined to be a repeated defect.

[0074] Output the feature information of repeated defects to determine the subsequent process based on whether their features meet the preset threshold parameters.

[0075] See also Figure 7 The present application provides an embodiment of a device for detecting repeated defects of a display screen, comprising: The processing unit 701 is used to perform consistency preprocessing on the collected display screen image; Optionally, the processing unit 701 includes: Perform color channel merging or splitting processing on the collected display screen images; Perform geometric correction on the display screen image; Pre-process and filter the display screen image to remove noise and background interference; The preprocessed image is enhanced, and the enhancement processing includes shadow correction and brightness enhancement.

[0076] The first generating unit 702 is used to perform point line block mura defect extraction processing on the display screen image, and perform defect area segmentation extraction processing to generate point line block defect areas; The second generating unit 703 is used to perform blob analysis on the point line block defect area to generate defect feature information; The defect merging unit 704 merges the defect areas of the point, line and block according to the aspect ratio AR value of the defect feature information and the defect position information; Optionally, the defect merging unit 704 includes: Determine the linear defect area and the non-linear defect area in the point-line block defect area according to the aspect ratio AR value in the defect feature information; When the linear defect area is within the minimum circumscribed rectangular area of ​​the non-linear defect area, the Euclidean distance between the linear defect area and the non-linear defect area is calculated; When the linear defect region is not located within the minimum circumscribed rectangular region of the non-linear defect region, determining the orientation information of the linear defect region and the non-linear defect region; Calculate the Euclidean distance between the positions of the nearest corner points of the minimum circumscribed rectangle of the linear defect area in the non-linear defect area according to the orientation information; Merge analysis was performed using Euclidean distance; According to the analysis results, the point, line and block defect areas are merged and split to obtain the display screen image with connected domain labels, and the Blob information and feature information of the corresponding positions after the merger are generated.

[0077] The first calculation unit 705 is used to calculate the geometric position adjacency matching degree according to the adjacency matching degree and the Euclidean distance of the two point-line block defect areas on each two display screen images for a plurality of display screen images with the same ID; Optionally, the first computing unit 705 includes: For several display screen images with the same ID, determine the independent area area, the intersecting area area and the minimum circumscribed rectangular area of ​​the two defect areas on every two display screen images; Calculate the adjacency matching degree of the defect area according to the area of ​​the independent area, the area of ​​the intersecting area and the area of ​​the minimum circumscribed rectangle; Calculate the Euclidean distance of the defective area according to the coordinate information of the two defective areas on each two display screen images; The geometric position adjacency matching degree is calculated based on the defect area adjacency matching degree and the Euclidean distance of the defect area.

[0078] An analysis unit 706 is used to perform display screen over-detection analysis processing according to the geometric position adjacency matching degree; The second calculation unit 707 is used to calculate the geometric position relationship information and defect feature relationship information of two point-line block defect areas on every two display screen images for a plurality of display screen images with different IDs; The second calculation unit 707 includes: For several display screen images with different IDs, calculate the Euclidean distance information of two point-line block defect areas on every two display screen images; Defect feature relationship information is calculated based on the roundness calculation value, contrast calculation value and texture feature calculation value of the two dot-line block defect areas on every two display screen images.

[0079] The detection unit 708 is used to perform batch abnormality detection of display screens based on the geometric position relationship information and the defect feature relationship information.

[0080] See also Figure 8 The present application provides a device for detecting repeated defects of a display screen, comprising: Processor 801 , memory 802 , input-output unit 803 , and bus 804 .

[0081] The processor 801 is connected to the memory 802 , the input and output unit 803 , and the bus 804 .

[0082] The memory 802 stores a program, and the processor 801 calls the program to execute the following steps: Figure 1 , Figure 2 and Figure 3 , Figure 4 , Figure 5 and Figure 6 The method in .

[0083] The present application provides a computer-readable storage medium, on which a program is stored, and when the program is executed on a computer, the program performs the following steps: Figure 1 , Figure 2 and Figure 3 , Figure 4 , Figure 5 and Figure 6 The method in .

[0084] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working processes of the systems, devices and units described above can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0085] In the several embodiments provided in the present application, it should be understood that the disclosed systems, devices and methods can be implemented in other ways. For example, the device embodiments described above are only schematic. For example, the division of the units is only a logical function division. There may be other division methods in actual implementation, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be an indirect coupling or communication connection through some interfaces, devices or units, which can be electrical, mechanical or other forms.

[0086] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0087] In addition, each functional unit in each embodiment of the present application may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit. The above-mentioned integrated unit may be implemented in the form of hardware or in the form of software functional units.

[0088] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present application is essentially or the part that contributes to the prior art or all or part of the technical solution can be embodied in the form of a software product. The computer software product is stored in a storage medium, including several instructions to enable a computer device (which can be a personal computer, server, or network device, etc.) to execute all or part of the steps of the method described in each embodiment of the present application. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM, read-only memory), random access memory (RAM, random access memory), disk or optical disk and other media that can store program code.

Claims

1. A method for detecting repeated defects in a display screen, characterized in that: include: Perform consistency preprocessing on the collected display screen images; Performing point-line-block mura defect extraction processing on the display screen image, and performing defect area segmentation extraction processing to generate point-line-block defect areas; Performing blob analysis on the defective area of ​​the dotted line block to generate defect feature information; Merging the defect areas of the point-line block according to the aspect ratio AR value of the defect feature information and the defect position information; For several display screen images with the same ID, the geometric position adjacency matching degree is calculated based on the adjacency matching degree and Euclidean distance of the defect areas of two point-line blocks on each two display screen images; Display screen over-inspection analysis processing is performed according to the geometric position adjacency matching degree.

2. The method according to claim 1, characterized in that After performing display screen over-inspection analysis processing according to the geometric position adjacency matching degree, the method further includes: For several display screen images with different IDs, calculate the geometric position relationship information and defect feature relationship information of two point-line block defect areas on every two display screen images; Batch abnormality detection of display screens is performed based on the geometric position relationship information and the defect feature relationship information.

3. The method according to claim 2, characterized in that The geometric position relationship information is Euclidean distance information; For several display screen images with different IDs, the geometric position relationship information and defect feature relationship information of the two point-line block defect areas on every two display screen images are calculated, including: For several display screen images with different IDs, calculate the Euclidean distance information of two point-line block defect areas on every two display screen images; Defect feature relationship information is calculated based on the roundness calculation value, contrast calculation value and texture feature calculation value of the two dot-line block defect areas on every two display screen images.

4. The method according to any one of claims 1 to 3, characterized in that Merging the point-line-block defect regions according to the aspect ratio AR value of the defect feature information and the defect position information includes: Determine the linear defect region and the non-linear defect region in the point-line block defect region according to the aspect ratio AR value in the defect feature information; When the linear defect region is within the minimum circumscribed rectangular region of the non-linear defect region, calculating the Euclidean distance between the linear defect region and the non-linear defect region; When the linear defect region is not located within the minimum circumscribed rectangular region of the non-linear defect region, determining the orientation information of the linear defect region and the non-linear defect region; Calculate the Euclidean distance between the positions of the nearest corner points of the minimum circumscribed rectangle of the linear defect area in the non-linear defect area according to the orientation information; Merge analysis was performed using Euclidean distance; The point, line, and block defect areas are merged and split according to the analysis results, a display screen image with a connected domain label is obtained, and Blob information and feature information of the corresponding position after the merger are generated.

5. The method according to any one of claims 1 to 3, characterized in that Perform consistency preprocessing on the collected display screen images, including: Perform color channel merging or splitting processing on the collected display screen images; Performing geometric correction processing on the display screen image; Pre-processing and filtering the display screen image to remove noise and background interference; The preprocessed image is enhanced, and the enhancement process includes shadow correction and brightness enhancement.

6. The method according to any one of claims 1 to 3, characterized in that For several display screen images with the same ID, the geometric position adjacency matching degree is calculated based on the adjacency matching degree and Euclidean distance of the two point-line block defect areas on each two display screen images, including: For several display screen images with the same ID, determine the independent area area, the intersecting area area and the minimum circumscribed rectangular area of ​​the two defect areas on every two display screen images; Calculate the adjacency matching degree of the defective area according to the area of ​​the independent area, the area of ​​the intersecting area and the area of ​​the minimum circumscribed rectangle; Calculate the Euclidean distance of the defective area according to the coordinate information of the two defective areas on each two display screen images; The geometric position adjacency matching degree is calculated according to the defect area adjacency matching degree and the Euclidean distance of the defect area.

7. A device for detecting repeated defects of a display screen, characterized in that: include: A processing unit, used for performing consistency preprocessing on the collected display screen images; A first generating unit is used to perform a point line block mura defect extraction process on the display screen image, and perform a defect area segmentation extraction process to generate a point line block defect area; A second generating unit is used to perform blob analysis on the point line block defect area to generate defect feature information; A defect merging unit is configured to merge the defect areas of the point-line block according to the aspect ratio AR value and the defect position information of the defect feature information; The first calculation unit is used to calculate the geometric position adjacency matching degree according to the adjacency matching degree and the Euclidean distance of the defect areas of two point-line blocks on each two display screen images for a plurality of display screen images with the same ID; The analysis unit is used to perform display screen over-inspection analysis processing according to the geometric position adjacency matching degree.

8. The device according to claim 7, characterized in that After the calculation unit, the device further comprises: The second calculation unit is used to calculate the geometric position relationship information and defect feature relationship information of two point-line block defect areas on every two display screen images for a plurality of display screen images with different IDs; A detection unit is used to perform batch abnormality detection of display screens based on the geometric position relationship information and the defect feature relationship information.

9. A device for detecting repeated defects in a display screen, characterized in that: include: Processor, memory, input-output unit, and bus; The processor is connected to the memory, the input and output unit, and the bus; The memory stores a program, and the processor calls the program to execute the method according to any one of claims 1 to 6.

10. A computer-readable storage medium having a program stored thereon, wherein the program, when executed on a computer, performs the method according to any one of claims 1 to 6.

Images