Detection Method for Liquid Crystal Display Screen and Liquid Crystal Display Screen Detection Device

Abstract

The present application provides a method for detecting a liquid crystal display screen and a liquid crystal display screen detection device. The method includes: obtaining a to-be-detected image of the liquid crystal display screen and a preset template image, wherein the template image includes a plurality of image interception regions; based on a trained image detection model, intercepting a plurality of target region images in the to-be-detected image according to the plurality of image interception regions; based on a preset gradient calculation rule, performing gradient calculation on each of the target region images to obtain gradient information in each of the target region images; determining whether each of the target region images meets a preset condition according to the gradient information corresponding to each of the target region images; if the number of target region images that meet the preset condition is greater than or equal to a preset threshold, determining that the liquid crystal display screen passes the detection. The present application detects the lamination effect of the liquid crystal display screen through images, reducing the detection cost, improving the detection efficiency and accuracy.

Description

Technical Field

[0001] This application relates to the technical field of display screen quality detection, and particularly to a detection method and a detection device for liquid crystal display screens. Background Art

[0002] Currently, for the detection of liquid crystal display screen products, many detections are carried out by detecting the functionality of conductive particles in the liquid crystal display screen to determine whether the liquid crystal display screen meets the usage requirements. However, the process of detecting the functions of conductive particles in the liquid crystal display screen is relatively cumbersome, requires a long time, and has a high cost. Summary of the Invention

[0003] The main purpose of this application is to provide a detection method and a detection device for liquid crystal display screens, aiming to improve the detection convenience, detection efficiency, and accuracy of detection results of liquid crystal display screens.

[0004] In the first aspect, this application provides a detection method for a liquid crystal display screen. The detection method for the liquid crystal display screen includes the following steps:

[0005] Obtain a to-be-detected image of the liquid crystal display screen and a preset template image, where the template image contains a plurality of image extraction regions;

[0006] Based on a trained image detection model, intercept a plurality of target region images in the to-be-detected image according to the plurality of image extraction regions;

[0007] Based on a preset gradient calculation rule, perform gradient calculation on each of the target region images to obtain gradient information in each of the target region images;

[0008] Determine whether each of the target region images meets a preset condition according to the gradient information corresponding to each of the target region images;

[0009] If the number of target region images that meet the preset condition is greater than or equal to a preset threshold, determine that the liquid crystal display screen passes the detection.

[0010] In the second aspect, this application also provides a detection device for a liquid crystal display screen. The detection device for the liquid crystal display screen includes a processor, a memory, and a computer program stored on the memory and executable by the processor. When the computer program is executed by the processor, the steps of the detection method for the liquid crystal display screen as described above are implemented.

[0011] The present application provides a method for detecting a liquid crystal display screen and a liquid crystal display screen detection device. The present application obtains a to-be-detected image of the liquid crystal display screen and a preset template image, wherein the template image includes a plurality of image interception regions; based on a trained image detection model, several target region images are intercepted in the to-be-detected image according to the plurality of image interception regions; based on a preset gradient calculation rule, gradient calculation is performed on each of the target region images to obtain gradient information in each of the target region images; it is determined whether each of the target region images meets a preset condition according to the gradient information corresponding to each of the target region images; if the number of target region images that meet the preset condition is greater than or equal to a preset threshold, it is determined that the liquid crystal display screen passes the detection. BRIEF DESCRIPTION OF THE DRAWINGS

[0012] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings required for use in the description of the embodiments will be briefly introduced below. Obviously, the drawings in the following description are some embodiments of the present application. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0013] Figure 1 It is a flowchart showing a method for detecting a liquid crystal display screen provided by an embodiment of the present application;

[0014] Figure 2 It is a schematic diagram of the scene of a target region image provided by an embodiment of the present application;

[0015] Figure 3 It is a schematic diagram of the scene of the gradient information of a target region image provided by an embodiment of the present application;

[0016] Figure 4 It is a schematic block diagram of the structure of a liquid crystal display screen detection device according to an embodiment of the present application. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0017] The technical solutions in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are some, but not all, of the embodiments of the present application. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present application without creative efforts fall within the scope of protection of the present application.

[0018] The flowchart shown in the drawings is only an example, and does not necessarily include all the content and operations / steps, nor does it necessarily need to be executed in the described order. For example, some operations / steps can be decomposed, combined, or partially merged, so the actual execution order may be changed according to the actual situation.

[0019] An embodiment of the present application provides a method for detecting a liquid crystal display screen and a liquid crystal display screen detection device. Among them, the method for detecting the liquid crystal display screen can be applied to the liquid crystal display screen detection device, and the liquid crystal display screen detection device can be a detection device including an image acquisition device. In addition, the method for detecting the liquid crystal display screen can be applied to a terminal device, which can be communicatively connected to an image acquisition device, and the terminal device can be an electronic device such as a tablet computer, a notebook computer, or a desktop computer.

[0020] The following will describe in detail some embodiments of the present application with reference to the accompanying drawings. Without conflict, the following embodiments and the features in the embodiments can be combined with each other.

[0021] Please refer to Figure 1 , Figure 1 which is a schematic flowchart of a method for detecting a liquid crystal display screen provided by an embodiment of the present application.

[0022] As Figure 1 shown, the method for detecting the liquid crystal display screen includes steps S101 to S104.

[0023] Step S101, obtain a to-be-detected image of the liquid crystal display screen and a preset template image, where the template image includes a plurality of image capture regions.

[0024] Exemplarily, the to-be-detected image of the liquid crystal display screen to be detected can be obtained by an image acquisition device communicatively connected to the liquid crystal display screen detection device, or the to-be-detected image can be obtained by an image acquisition device provided on the liquid crystal display screen detection device.

[0025] It can be understood that a plurality of ACF (Anisotropic Conductive Film) conductive particles and other unique structures of the liquid crystal display screen can be captured in the to-be-detected image, so that the pressing state of the liquid crystal display screen can be detected through the state of the captured ACF conductive particles. For example, if the pressing meets the requirements, the liquid crystal display screen can be used for subsequent installation in the terminal; or if the pressing does not meet the requirements, problems such as the liquid crystal display screen being insensitive to control or having display defects may occur. Therefore, the pressing state of the liquid crystal display screen can be determined through the state of the ACF conductive particles, so as to determine whether the liquid crystal display screen passes the detection.

[0026] Exemplarily, the template image can be a preset image. It can be understood that the template image can be an image corresponding to a liquid crystal display screen that has passed the detection, and the template image can be an image with manually marked image capture regions. Therefore, the template image includes a plurality of image capture regions.

[0027] For example, multiple image cropping regions can be determined in the template image through manual annotation, model prediction, and manual verification. Among them, the image cropping regions are used to crop the regions where the ACF particles are located in the template image. Due to manual annotation or manual verification, the image cropping regions on the template image will be relatively accurate and can be used to determine the target region images of the image to be detected.

[0028] Exemplarily, in the annotation of the image cropping regions of the template image, a first landmark point and a second landmark point can be marked on the template image. Among them, the first landmark point can be located at one end of the template image, and the second landmark point can be located at the other end of the template image, so that the template image can be more easily matched with other images to be detected. Moreover, when the sizes of other images to be detected are different from the size of the template image, the corresponding image cropping regions can be adjusted so that the image cropping regions can match the images to be detected, thereby improving the accuracy during image cropping and enhancing the accuracy of liquid crystal display detection.

[0029] Step S102: Based on the trained image detection model, crop several target region images from the image to be detected according to the multiple image cropping regions.

[0030] Exemplarily, input the image to be detected and the template image into the trained image detection model, so that the image detection model can crop the image to be detected through the multiple image cropping regions on the template image to obtain several target region images. It can be understood that each target region image may contain multiple ACF conductive particles, and whether each target region image meets the requirements can be determined based on the ACF conductive particles contained in each target region image, so as to determine whether the liquid crystal display corresponding to the image to be detected can pass the detection.

[0031] Exemplarily, the image detection model can be a Mobilenet-Yolov4 model to predict and locate the image cropping regions of the image to be detected.

[0032] In some embodiments, the trained image detection model crops several target region images from the image to be detected according to the multiple image cropping regions, including: determining the positions of the image cropping regions on the image to be detected based on the region matching network of the image detection model; and based on the image cropping network of the image detection model, cropping the image to be detected according to the positions of the image cropping regions on the image to be detected to obtain several target region images.

[0033] Exemplarily, the image detection model further includes a region matching network and an image cropping network. Among them, the region matching network can determine the positions of multiple image cropping regions in the image to be detected, and the image cropping network can crop the image to be detected.

[0034] For example, the image to be detected and the template image are input into the region matching network. According to the multiple image cropping regions on the template image, the position of each image cropping region on the image to be detected is determined in the image to be detected, thereby completing the determination of the position of the image cropping region on the image to be detected.

[0035] After determining the position of the image cropping region on the image to be detected, the image to be detected including the image cropping region is input into the image cropping network. According to the position of each image cropping region on the image to be detected, the image to be detected is cropped, thereby obtaining multiple target region images, as Figure 2 shown Figure 2 is a schematic diagram of the scenario of a target region image provided by an embodiment of the present application. It can be understood that multiple target region images are arranged according to a certain rule, and the corresponding image to be detected can be obtained.

[0036] In some embodiments, determining the positions of the image cropping regions on the image to be detected by the region matching network of the image detection model includes: performing binary processing on the image to be detected based on the image binary processing layer of the region matching network to obtain the binary image to be detected; performing connected component calculation on the binary image to be detected based on the connected component calculation layer of the region matching network to determine several connected regions on the binary image to be detected; determining the positions of the image cropping regions on the image to be detected according to the several connected regions and the image cropping regions based on the connected region matching layer of the region matching network.

[0037] Exemplarily, the region matching network further includes an image binary processing layer, a connected component calculation layer, and a connected region matching layer. Among them, the image binary processing layer can perform binary processing on the image, the connected component calculation layer can perform connected component calculation on the binary processed image, and the connected region matching layer is used to determine the positions of several connected regions in the image.

[0038] For example, after the image to be detected is input into the region matching network, the binarization processing layer in the region matching network performs image binarization on the image to be detected. Among them, the binarization process can classify the pixel values in the image to be detected. For example, the pixel value of point A in the image to be detected is 200, the pixel value of point B is 100, and the pixel threshold is 130. Therefore, point A is represented by 1 for the pixel value of this point, while point B is represented by 0 for the pixel value of this point, thus completing the binarization of the image to be detected. It should be noted that the above points A and B and their corresponding pixel values are only for illustrative purposes and do not limit the specific process of the image binarization of this application.

[0039] After the binarization processing layer completes the binarization of the image to obtain the binarized image to be detected, the binarized image to be detected is input into the connected component calculation layer to perform connected component operations on the binarized image to be detected in the connected component calculation layer, so as to be able to determine several connected regions on the binarized image to be detected. Among them, there may be several ACF conductive particles in each connected region. When the distance between one ACF conductive particle and another ACF conductive particle is relatively far, or several bit pixels near one ACF conductive particle are not adjacent to several bit pixels near another ACF conductive particle, it can be considered that these two ACF conductive particles exist in two connected regions.

[0040] Among them, after determining the connected components, the relatively dark regions in the connected components can be filled using the region hole filling algorithm to obtain a complete region.

[0041] After determining several connected regions on the image to be detected, in the connected region matching layer, the positions of each image cropping region on the image to be detected are determined according to multiple connected regions and the image cropping region. It can be understood that if the size of the connected region is the same as or similar to the size of the image cropping region, and the position of the image cropping region on the template image is the same as or approximate to the position of the connected region on the image to be detected, it can be considered that the image cropping region coincides with the connected region, that is, the connected region can be used as the image cropping region, thus completing the positioning of the image cropping region.

[0042] For example, the size and shape threshold can be set according to the size of the connected region. For example, the length threshold is 3 - 5 pixel points, and the width threshold is 1 - 4 pixel points. It can be understood that the number of pixel points corresponding to the length threshold and the width threshold can be adjusted according to actual needs and is not limited here; when the length of an image cropping region meets the length threshold and the width meets the width threshold, it can be considered that the size of the image cropping region is the same as or similar to the size of the connected region, so as to be able to determine the positions of multiple image cropping regions on the image to be detected.

[0043] The image detection model can roughly locate the ACF conductive particles on the image to be detected, and intercept the image to be detected based on the rough location to obtain multiple target region images. Analyzing the ACF conductive particles in the target region images can improve the detection accuracy of the liquid crystal display screen.

[0044] In some embodiments, before intercepting several target region images in the image to be detected according to multiple image interception regions based on the trained image detection model, it further includes: scaling the image interception regions according to the image to be detected and the template image to obtain target interception regions; the step of intercepting several target region images in the image to be detected according to multiple image interception regions based on the trained image detection model includes: intercepting several target region images in the image to be detected according to multiple target interception regions based on the trained image detection model.

[0045] Exemplarily, during the detection of multiple liquid crystal display screens, since the sizes of the liquid crystal display screens or the sizes of the images to be detected corresponding to the obtained liquid crystal display screens are different, it may cause a large difference between the size of the image to be detected and the size of the template image. In this case, the image interception regions on the template image may not be applicable to the image to be detected, so it is necessary to scale the image interception regions.

[0046] Exemplarily, the deviation information between the image to be detected and the template image can be determined through the fiducial points on the image to be detected and the fiducial points on the template image, so as to scale the image interception regions on the template image. Among them, the deviation information includes size deviation and position deviation.

[0047] It can be understood that the determination of the fiducial points on the image to be detected / template image can be the marker points on the liquid crystal display screen, or the fiducial points can be determined according to the edges of the image to be detected / template image.

[0048] For example, when taking pictures of the image to be detected / template image, the entire liquid crystal display screen is photographed. The fiducial points of the image to be detected / template image can be determined through the marker points on the liquid crystal display screen, such as welding points.

[0049] In some embodiments, the step of scaling the image interception regions according to the image to be detected and the template image to obtain target interception regions includes: determining the first and second fiducial points on the image to be detected and the third and fourth fiducial points on the template image; determining the scaling ratio of the image interception regions according to the first fiducial point, the second fiducial point, the third fiducial point and the fourth fiducial point; performing a scaling process on the image interception regions according to the scaling ratio to obtain target interception regions.

[0050] Exemplarily, as described in the above steps, the first fiducial point and the second fiducial point are determined on the image to be detected, wherein the first fiducial point and the second fiducial point are oppositely arranged on both sides of the image to be detected. For example, the first fiducial point is arranged on the left side of the center line of the image to be detected, and the second fiducial point is arranged on the right side of the center line of the image to be detected. Similarly, the third fiducial point and the fourth fiducial point on the template image are determined, and the third fiducial point and the fourth fiducial point are set in the same way as the first fiducial point and the second fiducial point, which will not be elaborated here.

[0051] After determining the first, second, third, and fourth fiducial points, the scaling ratio of the image cropping region can be determined according to the first, second, third, and fourth fiducial points.

[0052] In some embodiments, determining the scaling ratio of the image cropping region according to the first fiducial point, the second fiducial point, the third fiducial point, and the fourth fiducial point includes: determining a first distance according to the coordinates of the first fiducial point and the coordinates of the second fiducial point; determining a second distance according to the coordinates of the third fiducial point and the coordinates of the fourth fiducial point; and calculating the scaling ratio of the image cropping region according to the first distance and the second distance based on a preset scaling ratio calculation rule.

[0053] Exemplarily, a coordinate system is established for the image to be detected to determine the first positional relationship between the first fiducial point and the second fiducial point; a coordinate system is established for the template image to determine the second positional relationship between the third fiducial point and the fourth fiducial point, and the scaling ratio is determined according to the first positional relationship and the second positional relationship. Specifically, the first positional relationship may be the difference between the x coordinate of the first fiducial point and the x coordinate of the second fiducial point, and the second positional relationship is the difference between the x coordinate of the third fiducial point and the x coordinate of the fourth fiducial point. The two differences are divided, and according to the result of the division, it is determined whether the image cropping region is enlarged or reduced, and the corresponding ratio of enlargement or reduction.

[0054] Exemplarily, after calculating the scaling ratio, the image cropping region is scaled to obtain a target cropping region, so that the image to be detected can be processed through the target cropping region. As described in the above steps, it will not be elaborated here.

[0055] Scaling the image cropping region through the fiducial points of the image to be detected and the template image can improve the accuracy of the image detection model in determining the target region image of the image to be detected, thereby improving the accuracy of liquid crystal display detection.

[0056] In some embodiments, training data can be obtained, and the image detection model is trained according to the training data to obtain a trained image detection model.

[0057] Exemplarily, the training data may include manually annotated sample template images and sample images to be detected. Among them, each sample template image contains multiple sample image cropping regions. The sample template image and the sample images to be detected are input into the image detection model, so that the image detection model determines the image cropping regions of the sample images to be detected. After the image cropping regions of the sample images to be detected are determined, the sample images to be detected with the determined image cropping regions are compared with the annotated sample template images, so as to determine whether the image cropping regions in the sample images to be detected with the determined image cropping regions are reasonably determined, and adjust the parameters of the image detection model, thereby improving the prediction accuracy of the image detection model.

[0058] Exemplarily, the sizes of the sample template images and the sample images to be detected can be 416×416. Since the particle strips in the images are basically small targets, increasing the resolution can improve the accuracy of small target detection. During the training process, the images can also be rotated, the contrast and brightness can be changed, and the lossy compression rate of the pictures can be changed to achieve higher prediction accuracy.

[0059] Through the trained image detection model, the accuracy of locating the image cropping regions in the images to be detected can be effectively improved, so as to more accurately crop the target region images from the images to be detected, thereby improving the accuracy of analyzing whether the liquid crystal display screen passes the detection based on the target region images.

[0060] Step S103: Based on a preset gradient calculation rule, calculate the gradients of the target region images to obtain the gradient information in the target region images.

[0061] Exemplarily, the detection of ACF conductive particles on the liquid crystal display screen is mainly to distinguish between weakly pressed ACF particles and normal ACF particles. The blasting effect of normal ACF conductive particles is better, and the prominent regions can be well distinguished. However, the prominent regions of weakly pressed ACF conductive particles are not obvious, and they do not have good conductive characteristics. Some very weak blasting particles will not be considered as particles. In view of the fact that the production line uses a specified direction of lighting, with the light source from right to left, resulting in the imaging of ACF conductive particles being brighter on the right and darker on the left, a linear filter can be used for filtering first, which can effectively suppress noise and smooth the image. Then, the gradient information in the x and y directions of the target region image is obtained to obtain the gradient information in the target region.

[0062] Figure 3 This is a schematic diagram of the scenario of the gradient information of a target region image provided by an embodiment of the present application. It is possible to determine whether the ACF conductive particles are normal particles or weakly pressed particles through the gradient information of the target region, so as to determine whether the target region image can meet the preset conditions.

[0063] Step S104: Determine whether each of the target region images meets a preset condition according to the gradient information corresponding to each of the target region images.

[0064] Exemplarily, after calculating the gradient information of the target region image, it is possible to determine whether the target region image meets the preset condition according to the gradient information corresponding to the target region image.

[0065] For example, the position of the ACF conductive particles in the target region image and the area of the region of the ACF conductive particles can be determined according to the gradient information corresponding to the target region image, so as to determine whether the target region image meets the preset condition according to the position of the ACF conductive particles in the target region image and / or the area of the region of the ACF conductive particles.

[0066] In some embodiments, the determining whether each of the target region images meets a preset condition according to the gradient information corresponding to each of the target region images includes: determining the maximum gradient value in the target region image according to the gradient information corresponding to the target region image; determining the number of target regions and / or the positions of the target regions in the target region image according to the maximum gradient value; and determining whether the target region image meets the preset condition according to the number of target regions and / or the positions of the target regions in the target region image.

[0067] Exemplarily, after obtaining the gradient information, the maximum gradient value in the target region image is determined. It can be understood that in the region where there are ACF conductive particles, the gradient information is very strong, and each protrusion position of the ACF conductive particles has a certain region size, and the center position of this region is a maximum gradient value. And when the maximum gradient value is greater than or equal to a preset gradient threshold, this region can be determined as a target region. It can be understood that the target region can correspond to the ACF conductive particles and the nearby regions, that is, the position where this maximum value appears can be regarded as the center of the ACF conductive particles, so as to determine the number and / or position of the target regions in the target region image for indicating the ACF conductive particles, and be able to determine whether the target region image meets the preset condition according to the number and / or position of the ACF conductive particles in the target region image.

[0068] In some embodiments, the determining whether the target region image meets the preset condition according to the number of target regions and / or the positions of the target regions in the target region image includes at least one of the following: if the number of target regions in the target region image is greater than or equal to a preset target region number threshold, determine that the target region image meets the preset condition.

[0069] Exemplarily, it is possible to determine whether the target region image meets the preset conditions based on the number of target regions contained in the target region image. Specifically, the target region is characterized by ACF conductive particles, and it is determined whether the target region image meets the preset conditions based on the number of ACF conductive particles contained in the target region image.

[0070] For example, when the number of ACF conductive particles in the target region image is greater than or equal to the preset particle number threshold, it is determined that the target region image meets the preset conditions; when the number of ACF conductive particles in the target region image is less than the preset particle number threshold, it is considered that the target region image does not meet the preset conditions.

[0071] Determine the target distance between at least two target regions based on the positions of the target regions in the target region image, and if the target distance is greater than or equal to the preset distance threshold, determine that the target region image meets the preset conditions.

[0072] Specifically, the target region is characterized by ACF conductive particles, and it is determined whether the target region image meets the preset conditions by calculating the target distance between at least two ACF conductive particles.

[0073] It can be understood that the target region can be the region where the leftmost ACF conductive particle is located in the target region image and the region where the rightmost ACF conductive particle is located in the target region image. The distance in the x direction between these two ACF conductive particles can be referred to as the width, and it is determined whether the width is greater than or equal to the preset distance threshold, thereby determining whether the target region image meets the preset conditions.

[0074] Exemplarily, it is also possible to determine the target distance between at least two particles based on the particle positions in the target region image and determine whether the target region image meets the preset conditions based on the target distance.

[0075] For example, when the target distance is greater than or equal to the preset distance threshold, it is determined that the target region image meets the preset conditions; when the target distance is less than the preset distance threshold, it is considered that the target region image does not meet the preset conditions.

[0076] Based on the pixel uniformity calculation rule, calculate the pixel uniformity of the target region image based on the target regions in the target region image, and if the pixel uniformity of the target region image is greater than or equal to the preset uniformity threshold, determine that the target region image meets the preset conditions.

[0077] Specifically, the pixel uniformity here is characterized by the uniformity of ACF conductive particles, and it is determined whether the target region image meets the preset conditions by calculating the uniformity of at least two ACF conductive particles.

[0078] Exemplarily, calculate the particle uniformity in the target region image according to the particle positions in the target region image, and determine whether the target region image meets a preset condition based on the particle uniformity.

[0079] Among them, the particle uniformity can be calculated by the following formula:

[0080]

[0081] Among them, r(X,Y) is the uniformity of the arrangement of ACF conductive particles in a certain target region image, Cov(X,Y) is used to indicate the covariance of the X coordinates or Y coordinates of all ACF conductive particles in a certain target region image, Var[X] is used to indicate the variance of the gradient on the X coordinate of an ACF conductive particle, and Var[Y] is used to indicate the variance of the gradient on the Y coordinate of an ACF conductive particle.

[0082] After calculating the pixel uniformity corresponding to the particles through the above formula, determine the corresponding pixel uniformity through each target region in the target region image, so as to determine the pixel uniformity of the target region image. And when the pixel uniformity of the target region image is greater than or equal to a preset uniformity threshold, it is determined that the target region image meets the preset condition; when the pixel uniformity of the target region image is less than the preset uniformity threshold, it is considered that the target region image does not meet the preset condition.

[0083] It can be understood that the above three methods for determining whether the target region image meets the preset condition can be used alternatively, or any two or three of them can be used together, and the present application does not limit this.

[0084] Determining whether the target region image meets the preset condition through the above preset condition can improve the accuracy of liquid crystal display screen detection.

[0085] In some embodiments, perform binarization processing on each of the target region images to obtain a foreground image and a background image corresponding to each of the target region images; based on the foreground image corresponding to each of the target region images, determine the feature vector of the target region in the target region image, where the feature vector includes at least one of the center coordinates of the target region, the area of the target region, the width and / or height of the region where the target region is located, and the perimeter of the target region contour; the determining whether each of the target region images meets the preset condition according to the gradient information corresponding to each of the target region images includes: determining whether each of the target region images meets the preset condition according to the gradient information corresponding to each of the target region images and the feature vector of the target region.

[0086] Exemplarily, it is also possible to calculate the feature vectors of multiple target regions in the target region image, where the target region can be characterized by ACF particles, so as to determine whether the target region image meets the preset condition.

[0087] For example, perform image binarization on the target region image to obtain the foreground image and the background image corresponding to the target region image. The specific process of the binarization process can be as described in the above steps and will not be elaborated here. And determine the feature vector of the particles in the target region image based on the foreground image corresponding to the target region image. For example, blob analysis can be used to determine the feature vector of the particles in the target region image in the foreground image corresponding to the target region image. The feature vector of the particles includes at least one of the center coordinates of the particles, the area of the particles, the width and / or height of the region where the particles are located, and the perimeter of the particle contour.

[0088] It can be understood that after obtaining the foreground image, there will be some regions with relatively high gradients / pixel values near the particles, and these regions can be considered as the regions where the particles are located. Thus, the particles have feature vectors such as area, width and / or height of the region where they are located, and the perimeter of the particle contour.

[0089] After determining the feature vector of the particles, it can be determined whether the target region image meets the condition according to the feature vector of the particles and the feature vector threshold. For example, if the area of a particle contains 4 pixel points and the area threshold is 2 pixel points - 6 pixel points, it can be determined that the particle meets the preset condition. Thus, it can be determined whether the target region image meets the preset condition by whether the particles meet the preset condition. It can be understood that other feature vectors can be determined whether they meet the preset condition in the same way as the feature vector of the particle area described above and will not be elaborated here.

[0090] By the feature vector of the particles and the gradient information of the target region image, the accuracy of determining whether the target region image meets the preset condition can be improved, thereby improving the accuracy of the detection of the liquid crystal display screen.

[0091] Step S105: If the number of target region images that meet the preset condition is greater than or equal to the preset threshold, determine that the liquid crystal display screen passes the detection.

[0092] Exemplarily, when the number of target region images that meet the preset condition corresponding to a certain image to be detected is greater than or equal to the preset threshold, it can be determined that the liquid crystal display screen corresponding to the image to be detected passes the detection. Therefore, the corresponding liquid crystal display screen can enter the next detection item and / or be used in the terminal.

[0093] The detection method of the liquid crystal display screen provided by the above embodiment obtains a to-be-detected image of the liquid crystal display screen and a preset template image, wherein the template image includes a plurality of image extraction regions; based on a trained image detection model, a plurality of target region images are extracted from the to-be-detected image according to the plurality of image extraction regions; based on a preset gradient calculation rule, gradient calculation is performed on each of the target region images to obtain gradient information in each of the target region images; it is determined whether each of the target region images meets a preset condition according to the gradient information corresponding to each of the target region images; if the number of target region images that meet the preset condition is greater than or equal to a preset threshold, it is determined that the liquid crystal display screen passes the detection. There is no need to detect the conductivity of the ACF conductive particles in the liquid crystal display screen. It is more cost-effective and efficient to determine whether there are particles that do not meet the requirements through image detection. At the same time, compared with the detection and processing of other liquid crystal display screens, the present application can also improve the accuracy of liquid crystal display screen detection.

[0094] Please refer to Figure 4 , Figure 4 , which is a schematic block diagram of the structure of a liquid crystal display screen detection device provided by an embodiment of the present application. The liquid crystal display screen detection device can be a server or a terminal.

[0095] As Figure 4 shown, the liquid crystal display screen detection device includes a processor, a memory, and a network interface connected through a system bus. Among them, the memory can include a storage medium and an internal memory.

[0096] The storage medium can store an operating system and a computer program. The computer program includes program instructions. When the program instructions are executed, the processor can execute any detection method of the liquid crystal display screen.

[0097] The processor is used to provide computing and control capabilities to support the operation of the entire liquid crystal display screen detection device.

[0098] The internal memory provides an environment for the operation of the computer program in the storage medium. When the computer program is executed by the processor, the processor can execute any detection method of the liquid crystal display screen.

[0099] The network interface is used for network communication, such as sending assigned tasks, etc. Those skilled in the art can understand that Figure 4 the structure shown in

[0100] It should be understood that the processor may be a Central Processing Unit (CPU), and the processor may also be other general-purpose processors, Digital Signal Processors (DSPs), Application Specific Integrated Circuits (ASICs), Field-Programmable Gate Arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. Among them, the general-purpose processor may be a microprocessor or the processor may also be any conventional processor, etc.

[0101] It should be noted that those skilled in the art can clearly understand that for the convenience and brevity of description, the specific working process of the above-described fraud identification can refer to the corresponding process in the embodiment of the detection method of the liquid crystal display screen described above, and will not be elaborated here.

[0102] The embodiment of the present application also provides a computer-readable storage medium, on which a computer program is stored. The computer program includes program instructions, and the method implemented when the program instructions are executed can refer to each embodiment of the detection method of the liquid crystal display screen of the present application.

[0103] Among them, the computer-readable storage medium may be the internal storage unit of the liquid crystal display screen detection device described in the foregoing embodiment, such as the hard disk or memory of the liquid crystal display screen detection device. The computer-readable storage medium may also be an external storage device of the liquid crystal display screen detection device, such as a plug-in hard disk, a Smart Media Card (SMC), a Secure Digital (SD) card, a Flash Card, etc. equipped on the liquid crystal display screen detection device.

[0104] It should be understood that the terms used in the specification of the present application are only for the purpose of describing specific embodiments and are not intended to limit the present application. As used in the specification of the present application and the appended claims, unless the context clearly indicates otherwise, the singular forms "a", "an" and "the" are intended to include the plural forms.

[0105] It should also be understood that the term "and / or" used in the specification and appended claims of this application refers to any combination and all possible combinations of one or more of the associated listed items, and includes these combinations. It should be noted that in this text, the term "comprises", "comprising" or any other variation thereof is intended to cover non-exclusive inclusion, so that a process, method, article or system comprising a series of elements not only includes those elements, but also includes other elements not expressly listed, or elements inherent to such process, method, article or system. Without further limitation, an element defined by the statement "comprising an..." does not exclude the presence of additional identical elements in the process, method, article or system comprising the element.

[0106] The serial numbers of the embodiments of the present application above are only for description and do not represent the superiority or inferiority of the embodiments. The above is only the specific implementation manner of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art within the technical scope disclosed by the present application can easily think of various equivalent modifications or substitutions, and these modifications or substitutions should be covered within the protection scope of the present application. Therefore, the protection scope of the present application shall be subject to the protection scope of the claims.

Claims

1. A method for detecting a liquid crystal display screen, characterized in that, Including: Obtain a to-be-detected image of a liquid crystal display screen and a preset template image, where the template image contains a plurality of image cropping regions; Based on a trained image detection model, crop a plurality of target region images from the to-be-detected image according to the plurality of image cropping regions; Based on a preset gradient calculation rule, perform gradient calculation on each of the target region images to obtain gradient information in each of the target region images; Determine whether each of the target region images meets a preset condition according to the gradient information corresponding to each of the target region images; If the number of target region images that meet the preset condition is greater than or equal to a preset threshold, determine that the liquid crystal display screen passes the detection.

2. The detection method of the liquid crystal display screen according to claim 1, characterized in that The step of cropping a plurality of target region images from the to-be-detected image according to the plurality of image cropping regions based on the trained image detection model includes: Based on the region matching network of the image detection model, determine the positions of the plurality of image cropping regions on the to-be-detected image; Based on the image cropping network of the image detection model, crop the to-be-detected image according to the positions of the plurality of image cropping regions on the to-be-detected image to obtain a plurality of target region images.

3. The detection method of the liquid crystal display screen according to claim 2, characterized in that, The step of determining the positions of the plurality of image cropping regions on the to-be-detected image based on the region matching network of the image detection model includes: Based on the image binarization processing layer of the region matching network, perform binarization processing on the to-be-detected image to obtain a binarized to-be-detected image; Based on the connected component calculation layer of the region matching network, perform connected component operation on the binarized to-be-detected image to determine a plurality of connected regions on the binarized to-be-detected image; Based on the connected region matching layer of the region matching network, determine the positions of the plurality of image cropping regions on the to-be-detected image according to the plurality of connected regions and the image cropping regions.

4. The detection method of the liquid crystal display screen according to any one of claims 1-3, characterized in that, The step of determining whether each of the target region images meets a preset condition according to the gradient information corresponding to each of the target region images includes: Determine the gradient maximum value in the target region image according to the gradient information corresponding to the target region image; Determine the number and / or position of target regions in the target region image according to the gradient maximum value, where a target region includes pixels with a gradient maximum value greater than or equal to a preset gradient threshold; Determine whether the target region image meets a preset condition according to the number and / or position of target regions in the target region image.

5. The detection method of the liquid crystal display screen according to claim 4, characterized in that, The step of determining whether the target region image meets a preset condition according to the number and / or position of target regions in the target region image includes at least one of the following: If the number of target regions in the target region image is greater than or equal to a preset target region number threshold, determine that the target region image meets a preset condition; Determine the target distance between at least two target regions according to the positions of the target regions in the target region image, and if the target distance is greater than or equal to a preset distance threshold, determine that the target region image meets a preset condition; Based on the pixel uniformity calculation rule, calculate the pixel uniformity of the target region image according to the target region in the target region image, and if the pixel uniformity of the target region image is greater than or equal to a preset uniformity threshold, determine that the target region image meets the preset conditions.

6. The detection method of a liquid crystal display screen according to any one of claims 1 to 3, characterized in that, The method further includes: Perform binarization processing on each of the target region images to obtain a foreground image and a background image corresponding to each of the target region images; Based on the foreground image corresponding to each of the target region images, determine the feature vector of the target region in the target region image, where the feature vector includes at least one of the center coordinates of the target region, the area of the target region, the width and / or height of the region where the target region is located, and the perimeter of the target region contour; The determining whether each of the target region images meets the preset conditions according to the gradient information corresponding to each of the target region images includes: Determine whether each of the target region images meets the preset conditions according to the gradient information corresponding to each of the target region images and the feature vector of the target region.

7. The detection method of the liquid crystal display screen according to any one of claims 1 to 3, characterized in that Before intercepting a plurality of target region images in the image to be detected according to the plurality of image intercepting regions based on the trained image detection model, it further includes: Scale the image intercepting region according to the image to be detected and the template image to obtain a target intercepting region; The intercepting a plurality of target region images in the image to be detected according to the plurality of image intercepting regions based on the trained image detection model includes: Based on the trained image detection model, intercept a plurality of target region images in the image to be detected according to the plurality of target intercepting regions.

8. The detection method of the liquid crystal display screen according to claim 7, characterized in that, The scaling the image intercepting region according to the image to be detected and the template image to obtain a target intercepting region includes: Determine the first and second landmark points on the image to be detected and the third and fourth landmark points on the template image; Determine the scaling ratio of the image intercepting region according to the first landmark point, the second landmark point, the third landmark point, and the fourth landmark point; Scale the image intercepting region according to the scaling ratio to obtain a target intercepting region.

9. The detection method of the liquid crystal display screen according to claim 8, characterized in that, The determining the scaling ratio of the image intercepting region according to the first landmark point, the second landmark point, the third landmark point, and the fourth landmark point includes: Determine the first distance according to the coordinates of the first landmark point and the coordinates of the second landmark point; Determine the second distance according to the coordinates of the third landmark point and the coordinates of the fourth landmark point; Based on a preset scaling ratio calculation rule, calculate the scaling ratio of the image intercepting region according to the first distance and the second distance.

10. A liquid crystal display detection device, characterized in that The liquid crystal display screen detection device includes a processor, a memory, and a computer program stored on the memory and executable by the processor, where when the computer program is executed by the processor, the steps of the liquid crystal display screen detection method according to any one of claims 1 to 7 are implemented.

Images