Panel edge defect detection method and apparatus, and device and storage medium

Abstract

The present application relates to the technical field of defect detection. Provided are a panel edge defect detection method and apparatus, and a device and a storage medium, which are used for solving the problem of the stability and efficiency of defect detection for liquid crystal panels being relatively low. The method comprises: performing image collection on a liquid crystal panel to be subjected to detection, so as to acquire a collected panel image; acquiring a plurality of regions-of-interest from the collected panel image, wherein the regions-of-interest comprise edge regions; acquiring fitting edge straight lines respectively corresponding to a plurality of edge regions; and for any edge region, determining, on the basis of the distance between each edge point of the edge region and a fitting edge straight line of the edge region, whether the edge region has a defect. That is, whether an edge contour of a liquid crystal panel has a defect can be determined by means of "quantitative measurement", so as to greatly improve the stability and efficiency of defect detection for the liquid crystal panel.

Description

Panel edge defect detection method, device, equipment and storage medium

[0001] Cross-reference to related applications

[0002] The present application claims priority from the Chinese patent application No. 2024107190782 entitled "Panel edge defect detection method, device, equipment and storage medium" filed on June 5, 2024 with the China Patent Office, the content of which is incorporated herein by reference in its entirety. TECHNICAL FIELD

[0003] The present application relates to the technical field of defect detection, and provides a panel edge defect detection method, device, equipment and storage medium. BACKGROUND

[0004] It is known that in the traditional liquid crystal panel defect detection method, the "manual visual inspection" method is often used for defect detection, wherein the professional inspector mainly relies on "naked eye observation" to check whether the liquid crystal panel has defects such as bright spots, dark spots, bright lines, dark lines and color spots. However, due to the problems of human fatigue and subjectivity of this method, inconsistent results are prone to occur, thereby resulting in low stability and efficiency of liquid crystal panel defect detection.

[0005] Therefore, how to improve the stability and efficiency of liquid crystal panel defect detection is a problem to be solved at present. SUMMARY

[0006] The embodiments of the present application provide a panel edge defect detection method, device, equipment and storage medium, which are used to solve the problems of low stability and efficiency of liquid crystal panel defect detection.

[0007] In one aspect, a panel edge defect detection method is provided, the method comprising:

[0008] Image acquisition is performed on a liquid crystal panel to be detected to obtain a panel acquisition image;

[0009] A plurality of regions of interest are obtained from the panel acquisition image, wherein the regions of interest include edge regions;

[0010] A plurality of fitting edge straight lines corresponding to the edge regions are obtained;

[0011] For any edge region, whether the edge region has a defect is determined according to the distance between each edge point of the edge region and the fitting edge straight line of the edge region.

[0012] The application has the advantages that: for any edge region, whether the edge region has defects is determined according to the distance between each edge point of the edge region and the fitting edge line of the edge region, that is, whether the edge profile of the liquid crystal panel has defects is determined by means of quantitative measurement, so compared with the prior art in which defects are detected by artificial visual inspection, the application can obviously improve the stability and efficiency of the liquid crystal panel defect detection.

[0013] In an implementation manner, the step of acquiring the fitting edge line corresponding to each of the plurality of edge regions comprises:

[0014] For any edge region, edge points are extracted in a line scanning manner and at a certain interval to acquire a first edge point set of the edge region;

[0015] The first edge point set is subjected to outlier screening to acquire a second edge point set;

[0016] Each edge point in the second edge point set is subjected to fitting to acquire the fitting edge line corresponding to the edge region.

[0017] The application has the advantages that: since the line scanning manner is used to acquire the edge point set, the edge point set can be acquired more simply and quickly; in addition, since the edge point set is subjected to outlier screening, the fitting degree of the edge region can be improved to obtain a more accurate and reliable fitting edge line.

[0018] In an implementation manner, the step of determining, for any edge region, whether the edge region has defects according to the distance between each edge point of the edge region and the fitting edge line of the edge region comprises:

[0019] For the edge region, the distance between each edge point of the edge region and the fitting edge line of the edge region is calculated to obtain the distance corresponding to each edge point;

[0020] A plurality of edge points with a distance greater than a preset distance threshold are determined;

[0021] In the plurality of edge points, whether the number of continuous edge points is greater than a preset number threshold is determined;

[0022] If the number of continuous edge points is determined to be greater than the preset number threshold, it is determined that the edge region has defects.

[0023] The application has the beneficial effect that since the edge region is determined to have defects only when the distance is greater than the preset distance threshold and the number of continuous edge points is greater than the preset number threshold, the application can filter out a small number of continuous edge points to avoid false detection, thereby further improving the stability and efficiency of liquid crystal panel defect detection.

[0024] In an implementation manner, the step of determining whether the any edge region has defects according to the distance between each edge point of the any edge region and the fitted edge straight line of the any edge region includes:

[0025] For the any edge region, the distance between each edge point of the any edge region and the fitted edge straight line of the any edge region is calculated to obtain the distance corresponding to the each edge point;

[0026] A plurality of edge points with a distance greater than a preset distance threshold are determined.

[0027] A plurality of defect areas are obtained according to the plurality of edge points.

[0028] If any defect area is greater than a preset area threshold, it is determined that the any edge region has defects.

[0029] The application has the beneficial effect that since the edge region is determined to have defects only when the distance is greater than the preset distance threshold and the defect area is greater than the preset area threshold, the application can filter out defects with a small area to avoid false detection, thereby further improving the stability and efficiency of liquid crystal panel defect detection.

[0030] In an implementation manner, when the region of interest includes an edge grinding region, after the plurality of regions of interest are obtained from the panel, the method further includes:

[0031] For any edge grinding region, edge grinding points are extracted using a line scanning method and at a certain interval to obtain coordinate points entering and leaving the any edge grinding region, respectively.

[0032] The edge grinding width of the liquid crystal panel is determined according to the coordinate points entering and leaving the any edge grinding region.

[0033] The application has the beneficial effect that since the line scanning method is used to extract the edge grinding points, the edge grinding points can be obtained more simply and quickly, thereby the edge grinding width of the liquid crystal panel is determined more efficiently to accurately obtain the length and width information of the chamfer region.

[0034] In an implementation, when the region of interest includes a chamfer region, after acquiring a plurality of regions of interest from the panel image, the method further includes:

[0035] For any two edge regions adjacent to the liquid crystal panel, using line scanning to acquire a plurality of common points of the two edge regions;

[0036] Performing outlier filtering on the plurality of common points to acquire a chamfer point set of the chamfer region between the two edge regions;

[0037] According to the minimum circumscribed rectangle of the chamfer point set, acquiring the length-width information of the chamfer region.

[0038] The application has the following advantages: since line scanning is used to acquire common points of any two edge regions, the common points can be acquired more simply and quickly; in addition, since outlier filtering is also performed on the common points, the minimum circumscribed rectangle of the chamfer point set can be accurately determined to accurately acquire the length-width information of the chamfer region, thereby providing precise liquid crystal panel information for the user.

[0039] In an implementation, the step of acquiring a region of interest from the panel image includes:

[0040] Acquiring a pixel point set of the panel image;

[0041] According to the distance between any two pixel points in the pixel point set, arranging each pixel point in the pixel point set in ascending order to obtain a first sorting sequence;

[0042] According to the first sorting sequence, calculating the local density of each pixel point;

[0043] According to the local density of each pixel point, calculating the minimum distance of each pixel point;

[0044] For any pixel point, multiplying the local density and the minimum distance of the pixel point after normalization to obtain the weight of the pixel point;

[0045] According to the weight of each pixel point, determining a plurality of cluster center points;

[0046] According to the size of the local density, assigning the remaining pixel points in the first sorting sequence except the plurality of cluster center points to the cluster where the nearest cluster center point is located to acquire the plurality of regions of interest.

[0047] The application has the advantages that: since the improved weighted K-neighbor clustering algorithm is used, subjectivity and randomness caused by artificial decision of the clustering center point can be avoided, and the accuracy of determining the clustering center point can be greatly improved, so that the region of interest can be further accurately obtained.

[0048] In an implementation manner, the step of arranging each pixel point in the pixel point set in ascending order according to the distance between any two pixel points in the pixel point set to obtain a first sorting sequence comprises:

[0049] For any one pixel point in the pixel point set, distances between all pixel points in the pixel point set except the any one pixel point and the any one pixel point are calculated to obtain a plurality of first distances;

[0050] The plurality of first distances are added to obtain a total distance of the any one pixel point;

[0051] The each pixel point is arranged in ascending order according to the total distance of the each pixel point to obtain the first sorting sequence.

[0052] The application has the advantage that: since the each pixel point is arranged in ascending order according to the total distance of the each pixel point, a basic condition for subsequently more conveniently and quickly determining the clustering center point can be provided.

[0053] In an implementation manner, the step of calculating the minimum distance of each pixel point according to the local density of the each pixel point comprises:

[0054] For any one pixel point in the first sorting sequence, distances between the any one pixel point and all pixel points in the first sorting sequence having a higher local density than the any one pixel point are calculated to obtain a plurality of second distances;

[0055] A minimum value in the plurality of second distances is determined as the minimum distance of the any one pixel point.

[0056] The application has the advantage that: since the minimum distance between the pixel point and any other pixel point having a higher local density is determined as the minimum distance of the pixel point, a basic condition for subsequently determining a more accurate clustering center point can be provided.

[0057] In an implementation manner, the step of determining a plurality of clustering center points according to the weight of the each pixel point comprises:

[0058] The each pixel point is arranged in descending order according to the weight of the each pixel point to obtain a second sorting sequence;

[0059] According to the slope change trend corresponding to the second sorting sequence, a critical pixel point corresponding to a maximum slope change value is determined;

[0060] A plurality of pixel points before the critical pixel point in the second sorting sequence are determined as cluster center points.

[0061] The application has the beneficial effects that: since only the pixel points with high local density and small distance are considered as the cluster center points, the accuracy of obtaining the region of interest can be greatly improved. In addition, since the critical center points of the cluster are determined based on the weight slope change trend, not only clear standards for selecting the cluster center can be provided, but also more ideal clustering effects can be provided for the data set with uneven density, thereby further improving the accuracy of obtaining the region of interest.

[0062] In one aspect, a panel edge defect detection device is provided, and the device comprises:

[0063] An image acquisition unit is configured to acquire an image of a liquid crystal panel to be detected, and obtain a panel acquisition image.

[0064] A region of interest acquisition unit is configured to obtain a plurality of regions of interest from the panel acquisition image, wherein the region of interest includes an edge region.

[0065] A fitting edge line acquisition unit is configured to obtain a fitting edge line corresponding to each of the edge regions.

[0066] A defect determination unit is configured to determine whether any edge region has a defect according to a distance between each edge point of the any edge region and a fitting edge line of the any edge region.

[0067] In one aspect, an electronic device is provided, which comprises a memory, a processor, and a computer program stored in the memory and executable on the processor, and the processor implements any of the above methods when executing the computer program.

[0068] In one aspect, a computer storage medium is provided, which stores computer program instructions executable by a processor, and the computer program instructions implement any of the above methods when executed by the processor. BRIEF DESCRIPTION OF DRAWINGS

[0069] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the related art, the drawings needed in the embodiment or related art description will be briefly introduced. Obviously, the drawings in the following description are only embodiments of the present application, and those skilled in the art can obtain other drawings according to the provided drawings without creative labor.

[0070] FIG. 1 is a schematic diagram of an application scenario provided by an embodiment of the present application;

[0071] FIG. 2 is a schematic diagram of a flow of a panel edge defect detection method provided by an embodiment of the present application;

[0072] FIG. 3 is a schematic diagram of a panel edge defect detection device provided by an embodiment of the present application.

[0073] In the figure, 10 is a panel edge defect detection device, 101 is a processor, 102 is a memory, 103 is an I / O interface, 104 is a database, 30 is a panel edge defect detection device, 301 is an image acquisition unit, 302 is a region of interest acquisition unit, 303 is a fitted edge straight line acquisition unit, 304 is a defect determination unit, 305 is an edging width determination unit, and 306 is a chamfer length-width acquisition unit. DETAILED DESCRIPTION

[0074] To make the objectives, technical solutions and advantages of the present application clearer, the technical solutions in the embodiments of the present application will be described clearly and completely below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only some of the embodiments of the present application, but not all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative work fall within the scope of protection of the present application. The embodiments in the present application and the features in the embodiments can be combined with each other arbitrarily without conflict. Moreover, although a logical order is shown in the flowchart, in some cases, the steps shown or described can be performed in an order different from that here.

[0075] As known, in the traditional liquid crystal panel defect detection method, a "manual visual detection" method is usually used for defect detection, in which, mainly relying on professional inspectors to check whether there are defects such as bright spots, dark spots, bright lines, dark lines and color spots in the liquid crystal panel through "naked eye observation". However, due to the problems of human fatigue and subjectivity of this method, inconsistent results are prone to occur, thereby leading to the problems of low stability and efficiency of liquid crystal panel defect detection.

[0076] Based on this, the embodiment of the present application provides a panel edge defect detection method. In the method, first, an image of a liquid crystal panel to be detected can be collected to obtain a panel collection image. Then, a plurality of regions of interest can be obtained from the panel collection image. The region of interest includes an edge region. Next, a plurality of edge region fitting edge straight lines corresponding to the plurality of edge regions can be obtained. Finally, for any edge region, whether the edge region has a defect can be determined according to the distance between each edge point of the edge region and the fitting edge straight line of the edge region. Therefore, in the embodiment of the present application, for any edge region, whether the edge region has a defect is determined according to the distance between each edge point of the edge region and the fitting edge straight line of the edge region, that is, the edge profile of the liquid crystal panel is determined by a "quantitative measurement" method, so compared with the "manual visual inspection" method in the prior art, the stability and efficiency of the liquid crystal panel defect detection can be greatly improved.

[0077] After introducing the design idea of the embodiment of the present application, the application scenarios to which the technical solution of the embodiment of the present application can be applied will be briefly introduced. It should be noted that the following application scenarios are only used to illustrate the embodiment of the present application, but not to limit it. In the specific implementation process, the technical solution provided by the embodiment of the present application can be flexibly applied according to actual needs.

[0078] As shown in FIG. 1, it is an application scenario provided by the embodiment of the present application. The application scenario can include a panel edge defect detection device 10.

[0079] The panel edge defect detection device 10 can be used for defect detection of a liquid crystal panel, such as a personal computer (PC), a server, a laptop, and the like. The panel edge defect detection device 10 can include one or more processors 101, a memory 102, an I / O interface 103, and a database 104. Specifically, the processor 101 can be a central processing unit (CPU), or a digital processing unit, and the like. The memory 102 can be a volatile memory, such as a random-access memory (RAM); the memory 102 can also be a non-volatile memory, such as a read-only memory, a flash memory, a hard disk drive (HDD), or a solid-state drive (SSD); or the memory 102 can be any other medium capable of carrying or storing desired program codes in the form of instructions or data structures and capable of being accessed by a computer, but is not limited thereto. The memory 102 can be a combination of the above memories. The memory 102 can store part of program instructions of the panel edge defect detection method provided in the embodiments of the present application, and the program instructions can be executed by the processor 101 to implement the steps of the panel edge defect detection method provided in the embodiments of the present application, so as to solve the problems of low stability and low efficiency of liquid crystal panel defect detection. The database 104 can be used to store the panel acquisition image, the region of interest, the fitted edge straight line, the distance corresponding to each edge point, the first edge point set, and the second edge point set, and the like, involved in the scheme provided in the embodiments of the present application.

[0080] In the embodiments of the present application, the panel edge defect detection device 10 can acquire the panel acquisition image through the I / O interface 103, and then the processor 101 of the panel edge defect detection device 10 can improve the stability and efficiency of the liquid crystal panel defect detection according to the program instructions of the panel edge defect detection method provided in the embodiments of the present application in the memory 102. In addition, the panel acquisition image, the region of interest, the fitted edge straight line, the distance corresponding to each edge point, the first edge point set, and the second edge point set, and the like, can be stored in the database 104.

[0081] Of course, the method provided by the embodiments of the present application is not limited to the application scenario shown in FIG. 1, and can also be used in other possible application scenarios, which are not limited by the embodiments of the present application. The functions that can be achieved by the devices in the application scenario shown in FIG. 1 will be described together in subsequent method embodiments, and will not be described too much here. In the following, the method of the embodiments of the present application will be introduced with reference to the accompanying drawings.

[0082] As shown in FIG. 2, it is a flowchart of the panel edge defect detection method provided by the embodiments of the present application. The method can be executed by the panel edge defect detection device 10 in FIG. 1. Specifically, the flow of the method is introduced as follows.

[0083] Step 201: Image acquisition is performed on the liquid crystal panel to be detected to obtain a panel acquisition image.

[0084] In order to further improve the stability and efficiency of the liquid crystal panel defect detection, in the embodiments of the present application, the stability and efficiency of the liquid crystal panel defect detection can be improved by "quantitative measurement" on the acquisition image of the liquid crystal panel. Specifically, first, image acquisition can be performed on the liquid crystal panel to be detected to obtain a panel acquisition image.

[0085] Step 202: A plurality of regions of interest are obtained from the panel acquisition image.

[0086] In the embodiments of the present application, the region of interest can include an edge region.

[0087] Further, in order to improve the stability and efficiency of the liquid crystal panel defect detection while improving the accuracy of the liquid crystal panel defect detection, in the embodiments of the present application, after obtaining the panel acquisition image, a plurality of regions of interest can be obtained from the panel acquisition image, so that the accuracy of the liquid crystal panel defect detection can be further improved by subsequently performing targeted defect detection on each region of interest.

[0088] Step 203: A plurality of edge regions each corresponding to a fitted edge straight line are obtained.

[0089] In order to realize the edge contour defect detection of the liquid crystal panel, in the embodiments of the present application, after obtaining a plurality of regions of interest (edge regions), a plurality of edge regions each corresponding to a fitted edge straight line can be obtained to prepare for subsequently determining whether the edge contour of the liquid crystal panel is defective by "quantitative measurement".

[0090] Step 204: For any edge region, whether the edge region has a defect is determined according to the distance between each edge point of the edge region and the fitted edge straight line of the edge region.

[0091] In the embodiments of the present application, after the fitting edge straight line corresponding to each edge region is obtained, in order to greatly improve the stability and efficiency of the liquid crystal panel defect detection, for any edge region, whether the edge region has a defect can be determined according to the distance between each edge point of the edge region and the fitting edge straight line of the edge region.

[0092] Further, since for any edge region, whether the edge region has a defect is determined according to the distance between each edge point of the edge region and the fitting edge straight line of the edge region, that is, whether the edge profile of the liquid crystal panel has a defect is determined by means of "quantitative measurement", compared with the defect detection by "manual visual inspection" in the prior art, the present application can obviously greatly improve the stability and efficiency of the liquid crystal panel defect detection.

[0093] In a possible implementation, in order to further improve the stability and efficiency of the liquid crystal panel defect detection, in the embodiments of the present application, when the fitting edge straight line corresponding to each edge region is obtained, the remaining boundary points can be fitted to obtain the fitting edge straight line by screening outliers.

[0094] Specifically, for any edge region, first, the first edge point set of the edge region can be obtained by using a line scanning method with a certain interval and extracting edge points at a certain interval; then, the second edge point set can be obtained by screening outliers from the first edge point set; finally, the fitting edge straight line corresponding to the edge region can be obtained by fitting each edge point in the second edge point set.

[0095] Further, since the line scanning method is used to obtain the edge point set, the edge point set can be obtained more simply and quickly; in addition, since the edge point set is also screened for outliers, a more accurate and reliable fitting edge straight line can be obtained by improving the fitting degree of the edge region.

[0096] In a possible implementation, in order to further improve the stability and efficiency of the liquid crystal panel defect detection, in the embodiments of the present application, when for any edge region, whether the edge region has a defect is determined according to the distance between each edge point of the edge region and the fitting edge straight line of the edge region, the number of continuous edge points can be used to determine whether the corresponding edge region has a defect.

[0097] That is, for any edge region, first, the distance between each edge point of any edge region and the fitting edge straight line of any edge region can be calculated to obtain the distance corresponding to each edge point; then, a plurality of edge points with a distance greater than a preset distance threshold can be determined; next, in the plurality of edge points, whether the number of continuous edge points is greater than a preset number threshold can be determined; finally, if it is determined that the number of continuous edge points is greater than the preset number threshold, it can be determined that any edge region has a defect. Conversely, if it is determined that the number of continuous edge points is not greater than the preset number threshold, it can be determined that any edge region does not have a defect.

[0098] Further, since the edge region is only determined to have a defect when the distance is greater than the preset distance threshold and the number of continuous edge points is greater than the preset number threshold, the present application can filter out a small number of continuous edge points to avoid false detection, thereby further improving the stability and efficiency of liquid crystal panel defect detection.

[0099] In a possible implementation, in order to further improve the stability and efficiency of liquid crystal panel defect detection, in the embodiments of the present application, when determining whether any edge region has a defect according to the distance between each edge point of any edge region and the fitting edge straight line of any edge region for any edge region, the corresponding edge region can be determined to have a defect according to the "defect area".

[0100] That is, for any edge region, first, the distance between each edge point of any edge region and the fitting edge straight line of any edge region can be calculated to obtain the distance corresponding to each edge point; then, a plurality of edge points with a distance greater than a preset distance threshold can be determined; next, a plurality of defect areas can be obtained according to the plurality of edge points; finally, if it is determined that any defect area is greater than a preset area threshold, it can be determined that any edge region has a defect. Conversely, if it is determined that any defect area is not greater than the preset area threshold, it can be determined that any edge region does not have a defect.

[0101] Further, since the edge region is only determined to have a defect when the distance is greater than the preset distance threshold and the number of continuous edge points is greater than the preset number threshold, the present application can filter out a small number of continuous edge points to avoid false detection, thereby further improving the stability and efficiency of liquid crystal panel defect detection.

[0102] In order to further improve the stability and efficiency of liquid crystal panel defect detection, in the embodiments of the present application, after obtaining a plurality of regions of interest, different processing methods can be used to process different types of regions of interest (edged region, chamfer region, suspected defect region).

[0103] In a possible implementation, specifically, if the region of interest includes an edge grinding region, after a plurality of regions of interest are acquired from the panel image, for any edge grinding region, edge grinding points can be extracted using a line scanning method at certain intervals to acquire coordinate points entering and leaving any edge grinding region; then, the edge grinding width of the liquid crystal panel can be determined according to the coordinate points entering and leaving any edge grinding region.

[0104] Further, since the edge grinding points are extracted using the line scanning method, the edge grinding points can be acquired more simply and quickly, and thus the edge grinding width of the liquid crystal panel can be determined more efficiently to accurately acquire the length-width information of the chamfer region.

[0105] In addition, for the edge grinding region, the edge grinding region shape can be extracted according to the binarization result of the region of interest content, and since the edge grinding region is specularly reflected by the cutting surface of the edge grinding, the extracted edge grinding region is black as a whole.

[0106] In a possible implementation, if the region of interest includes a chamfer region (the chamfer region is located in the middle of two adjacent edge directions of the liquid crystal panel), after a plurality of regions of interest are acquired from the panel image, first, for any two adjacent edge regions of the liquid crystal panel, a plurality of common points of the two edge regions can be acquired by using a line scanning method on the two edge regions; then, the common points can be screened for outliers to acquire a chamfer point set of the chamfer region between the two edge regions; finally, the length-width information of the chamfer region can be acquired according to the minimum circumscribed rectangle of the chamfer point set.

[0107] Further, since the common points of the two edge regions are acquired using the line scanning method, the common points can be acquired more simply and quickly; in addition, since the common points are also screened for outliers, the length-width information of the chamfer region can be accurately acquired by accurately determining the minimum circumscribed rectangle of the chamfer point set, thereby providing the user with precise liquid crystal panel information.

[0108] In actual applications, the line scanning method can be horizontal straight line scanning and vertical straight line scanning, and both are scanning at certain intervals to record the coordinate point positions entering and leaving the edge grinding region / chamfer region, so as to measure the length-width information of the edge grinding region / chamfer region.

[0109] In a possible implementation, to further improve the stability and efficiency of the liquid crystal panel defect detection, in the embodiment of the present application, when the region of interest is acquired from the panel image, the region of interest can be acquired based on an improved density peak clustering algorithm based on weighted K-nearest neighbors.

[0110] Specifically, first, a pixel set of a panel acquisition image can be acquired; then, each pixel in the pixel set can be arranged in ascending order according to the distance between any two pixels in the pixel set to obtain a first sorting sequence; next, the local density of each pixel can be calculated according to the first sorting sequence; then, the minimum distance of each pixel can be calculated according to the local density of each pixel; next, for any pixel, the local density of any pixel can be multiplied after being normalized with the minimum distance to obtain the weight of any pixel; then, a plurality of clustering center points can be determined according to the weight of each pixel; finally, the remaining pixels in the first sorting sequence except for the plurality of clustering center points can be assigned to the cluster where the nearest clustering center point is located according to the size of the local density to obtain a plurality of regions of interest.

[0111] Further, since the improved weighted K-neighbor clustering algorithm is used, subjectivity and randomness caused by "manual decision of clustering center points" can be avoided, and the accuracy of determining the clustering center points can be greatly improved, so that the regions of interest can be further accurately obtained.

[0112] In a possible implementation, to further improve the stability and efficiency of liquid crystal panel defect detection, in the embodiment of the present application, when arranging each pixel in the pixel set in ascending order according to the distance between any two pixels in the pixel set to obtain a first sorting sequence, for any pixel in the pixel set, first, the distance between all pixels in the pixel set except the any pixel and the any pixel can be calculated to obtain a plurality of first distances; then, the plurality of first distances can be added to obtain the total distance of the any pixel; finally, each pixel can be arranged in ascending order according to the total distance of each pixel to obtain the first sorting sequence.

[0113] Further, since each pixel is arranged in ascending order according to the total distance of each pixel, a basic condition can be provided for more convenient and efficient determination of the clustering center points subsequently.

[0114] In a possible implementation, to further improve the stability and efficiency of liquid crystal panel defect detection, in the embodiment of the present application, when calculating the minimum distance of each pixel according to the local density of each pixel, for any pixel in the first sorting sequence, first, the distance between the any pixel and all pixels in the first sorting sequence whose local density is higher than that of the any pixel can be calculated to obtain a plurality of second distances; then, the minimum value in the plurality of second distances can be determined as the minimum distance of the any pixel.

[0115] Further, since the minimum distance between the pixel point and any other pixel point with higher local density is determined as the minimum distance of the pixel point, a basic condition can be provided for subsequent determination of more accurate cluster center points.

[0116] In a possible implementation, in order to further improve the stability and efficiency of the liquid crystal panel defect detection, in the embodiment of the present application, when the plurality of cluster center points are determined according to the weights of the respective pixel points, first, the respective pixel points can be arranged in descending order according to the weights of the respective pixel points to obtain a second sorting sequence; then, the critical pixel point corresponding to the maximum slope change can be determined according to the slope change trend corresponding to the second sorting sequence; finally, the plurality of pixel points before the critical pixel point in the second sorting sequence can be determined as the cluster center points.

[0117] Further, since only the pixel points with higher local density and smaller distance are considered as the cluster center points, the accuracy of obtaining the region of interest can be greatly improved. In addition, since the critical center point of the cluster is also determined based on the weight slope change trend, not only can a clear standard be provided for selecting the cluster center, but also a more ideal clustering effect can be provided for the data set with uneven density, thereby further improving the accuracy of obtaining the region of interest.

[0118] In a possible implementation, after the different types of regions of interest are processed respectively and the corresponding results are obtained, the results can be summarized to obtain different detection results. Of course, in order to facilitate the staff to process the edge defects in time, when it is determined that there is a real edge defect in the suspected defect region, the staff can be alarmed, and the corresponding panel acquisition image and the position coordinates of the edge defect and other information can be pushed to the staff.

[0119] To sum up, in the embodiment of the present application, for any edge region, whether the edge region has defects is determined according to the distance between the edge points of the edge region and the fitting edge straight line of the edge region, that is, whether the edge profile of the liquid crystal panel is defective is determined by "quantitative measurement", so compared with the "manual visual detection" in the prior art, the stability and efficiency of the liquid crystal panel defect detection can be greatly improved.

[0120] Based on the same inventive concept, the embodiment of the present application provides a panel edge defect detection device 30, as shown in FIG. 3, which comprises:

[0121] An image acquisition unit 301 is configured to acquire an image of a liquid crystal panel to be detected, and obtain a panel acquisition image.

[0122] The region-of-interest acquisition unit 302 is configured to acquire a plurality of regions of interest from the image captured by the panel, wherein the region of interest includes an edge region.

[0123] The fitted edge line acquisition unit 303 is configured to acquire a fitted edge line corresponding to each of the edge regions.

[0124] The defect determination unit 304 is configured to determine, for each of the edge regions, whether the edge region has a defect based on a distance between each edge point of the edge region and the fitted edge line of the edge region.

[0125] In an implementation, the fitted edge line acquisition unit 303 is further configured to:

[0126] extract the edge points of each of the edge regions using a line scanning method and at a certain interval to obtain a first edge point set of each of the edge regions;

[0127] perform outlier screening on the first edge point set to obtain a second edge point set;

[0128] fit each edge point in the second edge point set to obtain the fitted edge line corresponding to each of the edge regions.

[0129] In an implementation, the defect determination unit 304 is further configured to:

[0130] calculate the distance between each edge point of each of the edge regions and the fitted edge line of each of the edge regions to obtain a distance corresponding to each edge point;

[0131] determine a plurality of edge points with a distance greater than a preset distance threshold;

[0132] determine whether the number of continuous edge points in the plurality of edge points is greater than a preset number threshold;

[0133] if the number of continuous edge points is greater than the preset number threshold, determine that each of the edge regions has a defect.

[0134] In an implementation, the defect determination unit 304 is further configured to:

[0135] calculate the distance between each edge point of each of the edge regions and the fitted edge line of each of the edge regions to obtain a distance corresponding to each edge point;

[0136] determine a plurality of edge points with a distance greater than a preset distance threshold;

[0137] obtain a plurality of defect areas based on the plurality of edge points;

[0138] If it is determined that any of the defect areas is greater than the preset area threshold, it is determined that any of the edge regions has a defect.

[0139] In an implementation, the panel edge defect detection apparatus 30 further comprises an edge grinding width determination unit 305, configured to:

[0140] For any edge grinding region, edge grinding points are extracted using a line scanning manner and at intervals, and coordinate points entering and leaving any edge grinding region are obtained respectively.

[0141] According to the coordinate points entering and leaving any edge grinding region, the edge grinding width of the liquid crystal panel is determined.

[0142] In an implementation, the panel edge defect detection apparatus 30 further comprises a chamfer length-width obtaining unit 306, configured to:

[0143] For any two adjacent edge regions of the liquid crystal panel, a plurality of common points of the two edge regions are obtained by using a line scanning manner on the two edge regions.

[0144] The plurality of common points are subjected to outlier screening to obtain a chamfer point set of a chamfer region between the two edge regions.

[0145] According to a minimum circumscribed rectangle of the chamfer point set, length-width information of the chamfer region is obtained.

[0146] In an implementation, the region of interest obtaining unit 302 is further configured to:

[0147] Obtain a pixel point set of the panel acquisition image.

[0148] According to distances between any two pixel points in the pixel point set, each pixel point in the pixel point set is arranged in ascending order to obtain a first sorting sequence.

[0149] According to the first sorting sequence, local densities of the pixel points are calculated.

[0150] According to the local densities of the pixel points, minimum distances of the pixel points are calculated.

[0151] For any pixel point, the local density of the pixel point is multiplied by the minimum distance after normalization to obtain a weight of the pixel point.

[0152] According to the weights of the pixel points, a plurality of cluster center points are determined.

[0153] According to the local densities, the remaining pixel points in the first sorting sequence except the plurality of cluster center points are assigned to a cluster in which a nearest cluster center point is located, and a plurality of regions of interest are obtained.

[0154] In an implementation manner, the region of interest acquisition unit 302 is further configured to:

[0155] For any one of the pixel points in the pixel point set, distances between all the pixel points in the pixel point set and the any one pixel point are calculated, to obtain a plurality of first distances;

[0156] The plurality of first distances are added to obtain a total distance of the any one pixel point;

[0157] According to the total distance of each pixel point, the each pixel point is arranged in ascending order to obtain a first sorting sequence.

[0158] In an implementation manner, the region of interest acquisition unit 302 is further configured to:

[0159] For any one of the pixel points in the first sorting sequence, distances between the any one pixel point and all the pixel points in the first sorting sequence having a higher local density than the any one pixel point are calculated, to obtain a plurality of second distances;

[0160] A minimum value in the plurality of second distances is determined as a minimum distance of the any one pixel point.

[0161] In an implementation manner, the region of interest acquisition unit 302 is further configured to:

[0162] According to the weight of each pixel point, the each pixel point is arranged in descending order to obtain a second sorting sequence;

[0163] According to a slope change trend corresponding to the second sorting sequence, a critical pixel point corresponding to a maximum slope change value is determined;

[0164] A plurality of pixel points before the critical pixel point in the second sorting sequence are determined as cluster center points.

[0165] The panel edge defect detection device 30 can be used to execute the method executed in the embodiment shown in FIG. 2, and thus the functions and the like that can be achieved by each functional module of the panel edge defect detection device 30 can refer to the description of the embodiment shown in FIG. 2, and will not be described in detail.

[0166] In some possible implementation manners, each aspect of the method provided in the present application can also be implemented in the form of a program product, which includes program code. When the program product runs on a computer device, the program code is used to make the computer device execute the steps in the method according to various exemplary embodiments of the present application described in the specification, for example, the computer device can execute the method executed in the embodiment shown in FIG. 2.

[0167] Those skilled in the art can understand that all or part of the steps of the foregoing method embodiments can be completed by program instruction related hardware, and the foregoing program can be stored in a computer readable storage medium, and the program performs the steps of the foregoing method embodiments when executed; and the foregoing storage medium includes mobile storage equipment, read-only memory (ROM), random access memory (RAM), magnetic disc or optical disc, and various media that can store program codes. Alternatively, when the integrated units of the present application are realized in the form of software function modules and sold or used as independent products, they can also be stored in a computer readable storage medium. Based on this understanding, the technical solutions of the embodiments of the present application can be embodied in the form of software products, and the computer software products are stored in a storage medium, including a plurality of instructions for causing a computer device (which can be a personal computer, a server, or a network device, etc.) to execute all or part of the methods described in the embodiments of the present application. The foregoing storage medium includes mobile storage equipment, ROM, RAM, magnetic disc or optical disc, and various media that can store program codes.

[0168] Although the preferred embodiments of the present application have been described, those skilled in the art can make additional changes and modifications to the embodiments once they know the basic inventive concept. Therefore, the appended claims are intended to be interpreted as including all the preferred embodiments and all the changes and modifications falling within the scope of the present application.

[0169] Obviously, those skilled in the art can make various modifications and variations to the present application without departing from the spirit and scope of the present application. Thus, if these modifications and variations of the present application fall within the scope of the claims of the present application and their equivalents, the present application also intends to include these modifications and variations.

Claims

1. A method of panel edge defect detection, characterized by, The method comprises: image acquisition is performed on a liquid crystal panel to be detected to obtain a panel acquisition image; a plurality of regions of interest are obtained from the panel acquisition image; the regions of interest include edge regions; a fitting edge line corresponding to each of the edge regions is obtained; for any edge region, whether the edge region has a defect is determined according to distances between each edge point of the edge region and a fitting edge line of the edge region.

2. The method of claim 1, wherein, The step of obtaining a fitting edge line corresponding to each of the edge regions comprises: for any edge region, a first edge point set of the edge region is obtained by using a line scanning method and extracting edge points at certain intervals; outlier points in the first edge point set are screened to obtain a second edge point set; each edge point in the second edge point set is fitted to obtain a fitting edge line corresponding to the edge region.

3. The method of claim 1, wherein, The step of determining, for any edge region, whether the edge region has a defect according to distances between each edge point of the edge region and a fitting edge line of the edge region comprises: for any edge region, distances between each edge point of the edge region and a fitting edge line of the edge region are calculated to obtain distances corresponding to the edge points; a plurality of edge points with distances greater than a preset distance threshold are determined; whether the number of continuous edge points is greater than a preset number threshold is determined among the plurality of edge points; if the number of continuous edge points is determined to be greater than the preset number threshold, it is determined that the edge region has a defect.

4. The method of claim 1, wherein, The step of determining, for any edge region, whether the edge region has a defect according to distances between each edge point of the edge region and a fitting edge line of the edge region comprises: for any edge region, distances between each edge point of the edge region and a fitting edge line of the edge region are calculated to obtain distances corresponding to the edge points; a plurality of edge points with distances greater than a preset distance threshold are determined; a plurality of defect areas are obtained according to the plurality of edge points; if any defect area is determined to be greater than a preset area threshold, it is determined that the edge region has a defect.

5. The method of claim 1, wherein, When the regions of interest include edge regions, after the plurality of regions of interest are obtained from the panel acquisition image, the method further comprises: for any edge region, edge points are extracted using a line scanning method at certain intervals to obtain coordinate points entering and leaving the edge region, respectively; a beveling width of the liquid crystal panel is determined according to the coordinate points entering and leaving the edge region.

6. The method of claim 1, wherein, When the regions of interest include chamfer regions, after the plurality of regions of interest are obtained from the panel acquisition image, the method further comprises: for any two adjacent edge regions of the liquid crystal panel, a plurality of common points of the two edge regions are obtained by using a line scanning method. Outlier filtering is performed on the plurality of common points to obtain a chamfer point set of a chamfer region between the two edge regions; According to the minimum circumscribed rectangle of the chamfer point set, the length-width information of the chamfer region is obtained.

7. The method of claim 1, wherein, The step of obtaining the region of interest from the panel acquisition image comprises: Obtaining a pixel point set of the panel acquisition image; According to the distance between any two pixel points in the pixel point set, the pixel points in the pixel point set are arranged in ascending order to obtain a first sorting sequence; According to the local density of each pixel point, the minimum distance of each pixel point is calculated; For any pixel point, the local density and the minimum distance of the pixel point are multiplied after normalization to obtain the weight of the pixel point; According to the weight of each pixel point, a plurality of clustering center points are determined; According to the size of the local density, the remaining pixel points in the first sorting sequence except the plurality of clustering center points are assigned to the cluster where the nearest clustering center point is located to obtain the plurality of regions of interest. The step of arranging the pixel points in the pixel point set in ascending order to obtain a first sorting sequence according to the distance between any two pixel points in the pixel point set comprises:

8. The method of claim 7, wherein, For any pixel point in the pixel point set, the distance between all pixel points in the pixel point set except the pixel point and the pixel point is calculated to obtain a plurality of first distances; The plurality of first distances are added to obtain the total distance of the pixel point; According to the total distance of each pixel point, the pixel points are arranged in ascending order to obtain the first sorting sequence. The step of calculating the minimum distance of each pixel point according to the local density of each pixel point comprises:

9. The method of claim 7, wherein, For any pixel point in the first sorting sequence, the distance between the pixel point and all pixel points in the first sorting sequence with higher local density than the local density of the pixel point is calculated to obtain a plurality of second distances; The minimum value of the plurality of second distances is determined as the minimum distance of the pixel point. The step of determining a plurality of clustering center points according to the weight of each pixel point comprises:

10. The method of claim 7, wherein, According to the weight of each pixel point, the pixel points are arranged in descending order to obtain a second sorting sequence; According to the slope change trend corresponding to the second sorting sequence, a critical pixel point corresponding to the maximum slope change value is determined; The plurality of pixel points before the critical pixel point in the second sorting sequence are determined as the clustering center points. The device comprises:

11. A panel edge defect detection apparatus characterized by comprising: An image acquisition unit configured to acquire an image of a liquid crystal panel to be detected to obtain a panel acquisition image; A region of interest acquisition unit configured to obtain a plurality of regions of interest from the panel acquisition image; wherein the region of interest comprises an edge region; A fitting edge line acquisition unit configured to obtain a plurality of fitting edge lines corresponding to the edge regions, respectively; ​ The defect determining unit is configured to determine, for any one edge region, whether the any one edge region has a defect according to a distance between each edge point of the any one edge region and a fitting edge straight line of the any one edge region.

12. An electronic device, comprising: The device comprises: a memory for storing program instructions; a processor for invoking the program instructions stored in the memory and executing the method according to any one of claims 1-10 according to the obtained program instructions.

13. A storage medium, characterized by The storage medium stores computer executable instructions, and the computer executable instructions are used to make the computer execute the method according to any one of claims 1-10. The storage medium stores computer executable instructions, and the computer executable instructions are used to make the computer execute the method according to any one of claims 1-10.

Images