Image detection method and device, computer equipment and computer readable storage medium

Abstract

The invention discloses an image detection method and device, computer equipment and a computer readable storage medium, and aims to determine a target image detection result based on original image information and regional position information.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of computer, in particular to an image detection method and device, computer equipment and computer readable storage medium. BACKGROUND

[0002] In the manufacturing process of a display panel, key components of the display panel need to be detected for defects, for example, an electrostatic injury image of the display panel is collected by a camera, and then the display panel is detected for defect components based on the collected image. However, the current defect component positioning detection method based on the collected image has a single application scenario and cannot be applied to different display panels. SUMMARY

[0003] Embodiments of the present application provide an image detection method and device, computer equipment and computer readable storage medium.

[0004] In a first aspect, the present application provides an image detection method, comprising:

[0005] obtaining original image information;

[0006] performing region positioning on the original image information to obtain region position information;

[0007] determining a target image detection result based on the original image information and the region position information.

[0008] In a second aspect, the present application provides an image detection device, comprising:

[0009] an image acquisition module configured to obtain original image information;

[0010] a region positioning module configured to perform region positioning on the original image information to obtain region position information;

[0011] an image detection module configured to determine a target image detection result based on the original image information and the region position information.

[0012] In a third aspect, the present application further provides a computer equipment, comprising:

[0013] one or more processors;

[0014] a memory; and

[0015] one or more application programs, wherein the one or more application programs are stored in the memory and configured to be executed by the processor to implement the image detection method of any one of the first aspect.

[0016] In a fourth aspect, the embodiments of the present application provide a computer readable storage medium, which stores a computer program. The computer program is loaded by a processor to execute the steps in the image detection method according to any one of the first aspect. BRIEF DESCRIPTION OF DRAWINGS

[0017] In order to more clearly illustrate the technical solutions of the embodiments of the present application or the prior art, the drawings needed in the embodiments or the prior art description will be briefly introduced. Obviously, the drawings in the following description are only some embodiments described in the present application, and other drawings can be obtained by those skilled in the art without creative labor.

[0018] In order to more clearly illustrate the technical solutions of the embodiments of the present application or the prior art, the drawings needed in the embodiments or the prior art description will be briefly introduced. Obviously, the drawings in the following description are only some embodiments described in the present application, and other drawings can be obtained by those skilled in the art without creative labor.

[0019] Figure 1 is a scene schematic diagram of an image detection system provided by the embodiments of the present application;

[0020] Figure 2 is an embodiment flowchart of an image detection method provided by the embodiments of the present application;

[0021] Figure 3 is a schematic diagram of original image information provided by the embodiments of the present application;

[0022] Figure 4 is a specific embodiment flowchart of region positioning provided by the embodiments of the present application;

[0023] Figure 5 is Figure 3 a structural schematic diagram of a local region corresponding to the original image information of

[0024] Figure 6 is Figure 3 a structural schematic diagram of an intersection region in the original image information of

[0025] Figure 7 is a specific embodiment flowchart of image detection based on original image information and region position information provided by the embodiments of the present application;

[0026] Figure 8 is a principle block diagram of an image detection device provided by the embodiments of the present application;

[0027] Figure 9is an embodiment structure schematic diagram of the computer device provided by an embodiment of the present application. DETAILED DESCRIPTION

[0028] The technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments of the present application. Based on the embodiments in the present application, all other embodiments obtained by a person skilled in the art without creative work fall within the scope of protection of the present application.

[0029] In the description of the present application, it should be understood that the terms "center", "longitudinal", "lateral", "length", "width", "thickness", "upper", "lower", "front", "rear", "left", "right", "vertical", "horizontal", "top", "bottom", "inner", "outer" and the like indicate the orientation or positional relationship based on the orientation or positional relationship shown in the drawings, and are only for the convenience of describing the present application and simplifying the description, and do not indicate or imply that the device or element referred to must have a particular orientation, be constructed and operated in a particular orientation, and therefore cannot be understood as a limitation on the present application. In addition, the terms "first", "second", "third" and the like are only for descriptive purposes, and cannot be understood as indicating or implying relative importance or implicitly indicating the number of the technical features indicated. Therefore, the features defined with "first", "second", "third" and the like can explicitly or implicitly include one or more of the features. In the description of the present application, "several" means one or more than one, unless otherwise explicitly and specifically limited.

[0030] In the present application, the word "exemplary" is used to mean "serving as an example, instance, or illustration". Any implementation described as "exemplary" in the present application is not necessarily to be construed as preferred or advantageous over other implementations. The following description is presented to enable any person skilled in the art to make and use the present application. In the following description, for the purpose of explanation, details are set forth. It is apparent to those skilled in the art that the present application can be practiced without using these specific details. In other instances, well-known structures and processes are not described in detail in order to avoid obscuring the description of the present application. Thus, the present application is not intended to be limited to the embodiments shown, but is to be accorded the widest scope consistent with the principles and features presented herein.

[0031] It should be noted that the method of the embodiments of the present application is executed in the computer device, and the processing objects of each computer device exist in the form of data or information, such as time, which is essentially time information. It can be understood that if the size, quantity, position and the like are mentioned in subsequent embodiments, they all exist in the corresponding data for the computer device to process, and specific details are not described here.

[0032] The embodiment of the present application provides an image detection method and device, computer equipment and a computer readable storage medium, which are described in detail below.

[0033] Please refer to Figure 1 , Figure 1 The scene schematic diagram of the image detection system provided by the embodiment of the present application can include computer equipment 100, and the computer equipment 100 is integrated with an image detection device, such as Figure 1 computer equipment in the embodiment of the present application.

[0034] The computer equipment 100 in the embodiment of the present application is mainly used for acquiring original image information; performing region positioning on the original image information to obtain region position information; and determining a target image detection result based on the original image information and the region position information, so that accurate defect detection can be realized on display panels with different resolutions and sizes.

[0035] In the embodiment of the present application, the computer equipment 100 can be an independent server, or a server network or a server cluster composed of servers. For example, the computer equipment 100 described in the embodiment of the present application includes but is not limited to a computer, a network host, a single network server, a plurality of network server sets or a cloud server composed of a plurality of servers. The cloud server is composed of a large number of computers or network servers based on cloud computing.

[0036] It can be understood that the computer equipment 100 used in the embodiment of the present application can be a device including receiving and transmitting hardware, that is, a device having receiving and transmitting hardware capable of performing bidirectional communication on a bidirectional communication link. Such a device can include a cellular or other communication device with a single-line display or a multi-line display or a cellular or other communication device without a multi-line display. The specific computer equipment 100 can be a desktop terminal or a mobile terminal, and the computer equipment 100 can also be one of a mobile phone, a tablet computer, a notebook computer and the like.

[0037] Those skilled in the art can understand that the application environment shown in the embodiment of the present application is only one application scenario of the present application, and does not constitute a limitation on the application scenario of the present application. Other application environments can include more or fewer computer equipment than those shown in the embodiment of the present application. Figure 1 For example, only one computer equipment is shown in the embodiment of the present application, and it can be understood that the image detection system can also include one or more other services, which are not limited specifically herein. Figure 1 Figure 1 In addition, as shown in

[0038] In addition, as shown in Figure 1 ​As shown, the image detection system can further include a memory 200 for storing data, such as image data, for example, original image information, first filtered image information, and the like, such as gradient information, for example, first gradient information, second gradient information, and the like.

[0039] It should be noted that, Figure 1 The scene diagram of the image detection system shown is only an example, and the image detection system and the scene described in the embodiments of the present application are used to more clearly illustrate the technical solutions of the embodiments of the present application, and do not constitute a limitation on the technical solutions provided by the embodiments of the present application. Those skilled in the art can know that, as the image detection system evolves and new business scenarios appear, the technical solutions provided by the embodiments of the present application are also applicable to similar technical problems.

[0040] First, an image detection method is provided in the embodiments of the present application, and the execution subject of the image detection method is an image detection device. The image detection device is applied to a computer device, and the image detection method includes: obtaining original image information; performing region positioning on the original image information to obtain region position information; and determining a target image detection result based on the original image information and the region position information.

[0041] As Figure 2 As shown, the image detection method can include the following steps S201-S203, and the details are as follows.

[0042] S201, obtaining original image information.

[0043] The original image information is an original image of the display component collected by the imaging module. For example, the original image information can be an electrostatic discharge (ESD) grayscale image of the display component collected by the imaging module, or can be an original image of the display module captured by a high-definition camera, and the embodiments of the present application are not limited thereto. The image collected by the imaging module of other computer devices can also be obtained through a network, Bluetooth, infrared, and the like. For example, when the image detection method of the present application is applied to a smart phone, the smart phone can directly obtain the original image information through the imaging module configured by itself. When the image detection method of the present application is applied to a server, the server can collect the original image information through the imaging module of the smart phone, and obtain the original image information from the smart phone through a network, Bluetooth, infrared, and the like. Further, the display component described above represents a terminal medium for displaying image or video information, such as a display panel, a projection panel, and the like, and the embodiments of the present application are not limited thereto. For example, when the original image of the display component collected by the imaging module is an electrostatic discharge grayscale image of the display panel, since there is an insulating layer in the middle of the intersection line position of the display component, when static electricity accumulates, it will discharge and break through the insulating layer to cause the upper and lower lines to be conductive and cause display abnormalities, and the ESD grayscale image is a grayscale image for detecting whether the display panel has electrostatic discharge.

[0044] The original image information can include all display modules constituting the display component, or can include only part of the display component; further, the display component described above represents an independent functional unit having a specific function, such as a circuit board, a wire, and the like. Taking the grayscale image of the display panel as an example, the original image information includes the lines of the edge of the display panel, and when there is electrostatic discharge in the intersection area of the lines, the grayscale image will have defects in the intersection area. Therefore, based on the original image information, whether there is electrostatic discharge in the intersection area of the lines of the display panel and other panel defects can be detected, and the panel defect represents a characteristic representation of a pixel point that does not work normally in the image corresponding to the display panel, for example, a characteristic representation of a pixel point that does not work normally in the image corresponding to the display panel, such as uneven brightness, dead pixels, color abnormalities, and the like.

[0045] S202, region positioning is performed on the original image information to obtain region position information.

[0046] The display assembly includes at least two components, which can be metal components in any form. The at least two components can be in the same form or different forms, for example, the at least two components can be metal wires, or the at least two components can be metal sheets. The at least two components can be in regular shapes or irregular shapes, and the embodiments of the present application are not limited in this regard. The at least two components have an intersection region, which refers to a region where the at least two components intersect or cross each other at a point or a region. The intersection emphasizes the intersection point or region of the at least two components, which can be an intersection or crossing region on the same plane or different planes, and the present application is not limited in this regard.

[0047] The region position information is the position information of the intersection region of the at least two components in the original image information. For example, as shown in Figure 3 The display assembly includes three horizontal metal components (three white horizontal bars shown by the dashed box) and three vertical metal components (three white vertical bars shown by the dashed box). The position information of the intersection region of the horizontal metal components and the vertical metal components in the original image information is the region position information.

[0048] In a specific implementation, the original image information includes a plurality of first candidate regions and a plurality of second candidate regions. The plurality of first candidate regions are regions corresponding to the first component of the display assembly in the original image information, and the plurality of second candidate regions are regions corresponding to the second component of the display assembly in the original image information. The plurality of first candidate regions are regions extending along the horizontal direction, and the plurality of second candidate regions are regions extending along the vertical direction. For example, as shown in Figure 3 The plurality of first candidate regions are regions corresponding to the three white horizontal bars, and the plurality of second candidate regions are regions corresponding to the three white vertical bars.

[0049] It should be noted that in the embodiment, the plurality of first candidate regions are regions extending along the horizontal direction, and the plurality of second candidate regions are regions extending along the vertical direction. The intersection region is a region where the plurality of first candidate regions and the plurality of second candidate regions overlap. However, in other embodiments, the intersection region can be a region intersecting at a non-right angle, a region where the components intersect at a non-straight line, and the like, and the embodiments of the present application are not limited in this regard.

[0050] In a specific embodiment, as shown in Figure 4 The region positioning on the original image information in the step S202 to obtain the region position information can include the following steps S301-S303, and the details are as follows.

[0051] 301. Determine a first target region from the plurality of first candidate regions based on first gradient information corresponding to the first candidate regions; the first target region is a first candidate region with first gradient information greater than a preset first threshold.

[0052] The first gradient information in the embodiment is pixel gradient information of all pixel points in the first candidate region, and the first gradient information includes horizontal gradient information and / or vertical gradient information. The horizontal gradient information is determined based on pixel difference values of horizontally adjacent pixel points in the first candidate region, i.e., pixel points in the x direction as shown. Figure 3 The vertical gradient information is determined based on pixel difference values of vertically adjacent pixel points in the first candidate region, i.e., pixel points in the y direction as shown. Figure 3

[0053] In a specific embodiment, the first gradient information includes the horizontal gradient information, and the first gradient information of the first candidate region is obtained based on the following manner: for any first candidate region in the first candidate regions, calculating pixel difference values of horizontally adjacent pixel points in the first candidate region; summing the pixel difference values of the horizontally adjacent pixel points to obtain a first pixel accumulation value; and dividing the first pixel accumulation value by the number of pixels in the first candidate region to obtain the horizontal gradient information of the first candidate region. The calculation process of the pixel difference values of the horizontally adjacent pixel points can be represented as: dx = abs(int(blur_image[r][w]) - int(blur_image[r][w+1])), where dx represents the pixel difference values of the horizontally adjacent pixel points, blur_image[r][w] represents the pixel value of the pixel point in the rth row and the wth column, blur_image[r][w+1]) represents the pixel value of the pixel point in the rth row and the w+1th column, abs() represents the absolute value, and int() represents the integer. That is, the pixel difference values of the horizontally adjacent pixel points refer to the absolute values of the pixel difference values of the horizontally adjacent pixel points in the first candidate region.

[0054] ​In another specific embodiment, the first gradient information comprises longitudinal gradient information, and the first gradient information of the first candidate region is obtained based on the following manner: for any first candidate region of the first candidate regions, a pixel difference value of longitudinally adjacent pixels in the first candidate region is calculated; the pixel difference values of the longitudinally adjacent pixels are summed to obtain a second pixel accumulation value; and the second pixel accumulation value is divided by the number of pixels in the first candidate region to obtain the longitudinal gradient information of the first candidate region. The calculation process of the pixel difference value of the longitudinally adjacent pixels can be represented as: dy = abs(int(blur_image[r][w]) - int(blur_image[r+1][w])), where dy represents the pixel difference value of the longitudinally adjacent pixels, blur_image[r][w] represents the pixel value of the pixel at the rth row and the wth column, blur_image[r+1][w] represents the pixel value of the pixel at the (r+1)th row and the wth column, abs() represents the absolute value, and int() represents the integer. That is, the pixel difference value of the longitudinally adjacent pixels refers to the absolute value of the pixel difference value of the longitudinally adjacent pixels in the first candidate region.

[0055] The first target region is a candidate region of the first candidate regions in which the intersection region exists, and the preset first threshold is a gradient threshold value for measuring whether the first candidate regions exist the intersection region, which can be set according to actual needs. For example, the preset first threshold is set to a value close to 0.

[0056] Generally, the intersection region has obvious color difference relative to the non-intersection region, the gradient information of the first candidate region in which the intersection region does not exist is close to 0, and the gradient information of the first candidate region in which the intersection region exists is a larger value. Therefore, based on the first gradient information of the first candidate regions and the preset first threshold, the first target region in which the intersection region exists can be screened from the first candidate regions. For example, as shown in FIG. 5, the region in which the white horizontal bar intersects with the white vertical bar has obvious color difference relative to other regions, and therefore, the white horizontal bar in which the key points 5 and 6 are located has a larger region gradient relative to the white horizontal bar in which the key points 1 and 2 are located and the white horizontal bar in which the key points 3 and 4 are located. Thus, the white horizontal bar in which the key points 5 and 6 are located is the first target region. Figure 5

[0057] 302、based on the second gradient information corresponding to the second candidate region, determining a second target region from the second candidate regions; the second target region is a second candidate region in which the second gradient information is greater than a preset second threshold.

[0058] ​In a specific embodiment, the calculation of the second gradient information of the second candidate region is the same as the calculation of the first gradient information of the aforementioned first candidate region, and for brevity, the calculation of the first gradient information of the aforementioned first candidate region is referred to here.

[0059] The second target region is a candidate region in which the intersection region exists in the plurality of second candidate regions, and the preset second threshold is a gradient threshold value preset for measuring whether the intersection region exists in the plurality of second candidate regions. The preset second threshold can be set according to actual needs, for example, the preset second threshold is set to a value close to 0.

[0060] Generally, the intersection region has a significant color difference relative to the non-intersection region, the gradient information of the second candidate region in which the intersection region does not exist is close to 0, and the gradient information of the second candidate region in which the intersection region exists is a larger value. Therefore, based on the second gradient information of the plurality of second candidate regions and the preset second threshold, the second target region in which the intersection region exists can be screened from the plurality of second candidate regions. For example, continuing to refer to FIG. 7, the white vertical bar in which the key point 7 and the key point 8 are located, the white vertical bar in which the key point 9 and the key point 10 are located, and the white vertical bar in which the key point 11 and the key point 12 are located all have the intersection with the white horizontal bar in which the key point 5 and the key point 6 are located, and the gradient information of the white vertical bar in which the key point 7 and the key point 8 are located, the white vertical bar in which the key point 9 and the key point 10 are located, and the white vertical bar in which the key point 11 and the key point 12 are located is a larger value. Therefore, based on the second gradient information and the preset second threshold, the white vertical bar in which the key point 7 and the key point 8 are located, the white vertical bar in which the key point 9 and the key point 10 are located, and the white vertical bar in which the key point 11 and the key point 12 are located can be determined as the second target region. Figure 5

[0061] 303、Based on the first target region and the second target region, determine the region position information.

[0062] The region position information is the region position information of the intersection region of the first target region and the second target region. After the first target region in which the intersection region exists and the second target region in which the intersection region exists are determined, the position information of the intersection region can be determined based on the first target region and the second target region.

[0063] In a specific embodiment, the step of determining the region position information based on the first target region and the second target region can specifically include: determining the first position information based on the first target region; determining the second position information based on the second target region; and determining the region position information based on the first position information and the second position information.

[0064] For example, continuing to refer to FIG. 7, the first position information is the position information corresponding to the edge point of the white horizontal bar in which the key point 5 and the key point 6 are located, and the second position information is the position information corresponding to the edge point of the white vertical bar in which the key point 7 and the key point 8 are located.​Figure 5 As shown, the first target area is the white horizontal bar where key points 5 and 6 are located, and the first location information is the location information corresponding to key points 5 and 6.

[0065] The second location information is the location information corresponding to the edge points of the second target area. For example, continuing to refer to... Figure 5 As shown, the second target area includes the white vertical bars where key points 7 and 8 are located, the white vertical bars where key points 9 and 10 are located, and the white vertical bars where key points 11 and 12 are located. The second location information includes the location information corresponding to key points 7, 8, 9, 10, 11 and 12.

[0066] Based on the first and second location information, the location information of the intersecting regions of intersecting components in the original image information can be determined. For example, combining... Figure 5 and Figure 6 As shown, based on the first location information and the second location information, it can be determined that... Figure 6 The three intersecting regions shown from left to right are: the leftmost intersecting region is determined based on the positional information of key points 5, 6, 7 and 8; the middle intersecting region is determined based on the positional information of key points 5, 6, 9 and 10; and the rightmost intersecting region is determined based on the positional information of key points 5, 6, 11 and 12.

[0067] In one specific embodiment, before performing region localization on the original image information to obtain region location information, the original image information is updated based on the following method:

[0068] (1) Perform the first filtering process on the original image information to obtain the first filtered image information.

[0069] The first filtering process performed on the original image information in this embodiment can reduce noise interference in the original image information and improve the accuracy of image detection results. The method for the first filtering process on the original image information in this embodiment can adopt existing filtering enhancement methods, such as box filtering, mean filtering, Gaussian filtering, median filtering, bilateral filtering, etc. For details, please refer to existing filtering enhancement methods, which will not be elaborated here.

[0070] (2) Perform a second filtering process on the first filtered image information to obtain new original image information.

[0071] The second filtering processing on the first filtered image information can obtain the region contour information of the region where the component is located in the original image information and filter out the region irrelevant to the component, thereby improving the accuracy of the image detection result. In a specific implementation manner, the second filtering processing on the first filtered image information to obtain new original image information includes: performing threshold segmentation on the first filtered image information to obtain region contour information; performing connected domain detection on the region contour information to obtain connected domain information; and filtering the region contour information based on the connected domain information to obtain the new original image information.

[0072] Threshold segmentation is a region-based image segmentation technique, which divides image pixels into several categories by setting different feature thresholds. The region contour information is the region contour information of the region where the component is located in the original image information obtained by threshold segmentation on the first filtered image information. For example, as shown in FIG. 6, the region contour information includes the contour information of the region where the white horizontal bar is located and the contour information of the region where the white vertical bar is located. Figure 3

[0073] In a specific embodiment, the step of performing threshold segmentation on the first filtered image information to obtain the region contour information can specifically include: performing threshold segmentation on the first filtered image information to obtain a binary image; and determining the region contour information based on the binary image. The binary image refers to an image in which the pixel value of a pixel point in a foreground region (i.e., the region where the component is located in the original image information) is a first pixel value and the pixel value of a pixel point in a background region (i.e., a region other than the region where the component is located in the original image information) is a second pixel value. The pixel value refers to the unit of resolution in the horizontal and vertical directions of an image information, which can represent the size of a small dot or a pixel in an image information. For example, the binary image refers to an image in which the pixel value of a pixel point in a foreground region is 255 and the pixel value of a pixel point in a background region is 0. By performing threshold segmentation on the first filtered image information, the region where the component is located in the original image information can be distinguished from the background region, thereby facilitating the positioning of the region where the component is located in the original image information.

[0074] Further, the step of performing threshold segmentation on the first filtered image information to obtain a binary image can specifically include: comparing, for each pixel point in the first filtered image information, the pixel value of the pixel point with a preset pixel threshold value; if the pixel value of the pixel point is greater than the pixel threshold value, setting the pixel value of the pixel point as a first pixel value; and if the pixel value of the pixel point is less than or equal to the pixel threshold value, setting the pixel value of the pixel point as a second pixel value.

[0075] ​In a specific embodiment, the connected domain information is information of connected domains in the region contour information obtained by performing connected domain detection on the region contour information, and a connected domain refers to a region composed of pixels with the same pixel value and adjacent positions in the region contour information. In the embodiment of the present application, the connected domain in which the component in the original image is located usually has a large connected domain size. By performing connected domain detection on the region contour information, the connected domain information in the region contour information is obtained, and then the region contour information is filtered based on the connected domain information, so that the interference region with a small connected domain size can be filtered out from the region contour information, thereby improving the accuracy of the image detection result.

[0076] In the embodiment of the present application, the connected domain size includes but is not limited to the connected domain area, the length of the circumscribed rectangle of the connected domain, and the width of the circumscribed rectangle of the connected domain, and the embodiment of the present application is not limited thereto. In a specific embodiment, the connected domain size includes the connected domain area, the length of the circumscribed rectangle of the connected domain, and the width of the circumscribed rectangle of the connected domain, and the preset size threshold includes a first size threshold, a second size threshold, and a third size threshold. The above filtering of the region contour information based on the connected domain information to obtain the second filtered image information includes: for each connected domain in the region contour information, obtaining the connected domain area, the length of the circumscribed rectangle, and the width of the circumscribed rectangle of the connected domain from the connected domain information; determining a target connected domain from the multiple connected domains based on the connected domain area, the length of the circumscribed rectangle, and the width of the circumscribed rectangle; and filtering the region contour information based on the target connected domain to obtain the second filtered image information.

[0077] When the target connected domain is determined from the multiple connected domains based on the connected domain area, the length of the circumscribed rectangle, and the width of the circumscribed rectangle, the connected domain area, the length of the circumscribed rectangle, and the width of the circumscribed rectangle can be compared with the first size threshold, the second size threshold, and the third size threshold, respectively. If the connected domain area is less than the first size threshold, and / or the length of the circumscribed rectangle is less than the second size threshold, and / or the width of the circumscribed rectangle is less than the third size threshold, the connected domain is determined as the target connected domain.

[0078] S203, determining a target image detection result based on the original image information and the region position information.

[0079] The target image detection result is an object detection result corresponding to the intersection region of the intersecting components in the original image information. For example, the target image detection result is a defect detection result corresponding to the intersection region of the intersecting components in the original image information. The target image detection result includes two kinds of intersecting component presence objects and intersecting component absence objects, for example, the target image detection result includes two kinds of intersecting component presence defects and intersecting component absence defects. The embodiment determines the image detection result based on the original image information and the region position information, which can achieve accurate defect detection at different resolutions and sizes.

[0080] In a specific implementation, referring to Figure 7 The step of determining the target image detection result based on the original image information and the region position information in the step S203 can include steps S401-S402, and details are as follows.

[0081] The step S401 inputs the original image information into an object detection model to output object position information of the original image information through the object detection model.

[0082] The object position information is position information corresponding to an object in the original image information. The object can be any target object in the original image information, for example, the object is a defect in the original image information, and the object position information is position information corresponding to the defect in the original image information. The object detection model is a pre-trained network model for object detection of the original image information. The original image information is input into the object detection model, and the object detection model can output the object position information of the original image information. For example, the object detection model is a pre-trained network model for defect detection of the original image information. The original image information is input into the object detection model to obtain defect position information of the original image information. Based on the defect position information, it can be determined which positions in the original image information have defects.

[0083] In an embodiment, the object detection model adopts a memory-based end-to-end segmentation network (MemSeg) model. The MemSeg model is based on a U-Net architecture, uses a pre-trained ResNet18 as an encoder, and introduces an analog abnormal sample and a memory module to assist model learning in a more guided manner, thereby completing a semi-supervised object detection task in an end-to-end manner, solving the problems of sample imbalance of data in a complex scene and difficulty in collecting defect data, and improving the robustness of the model.

[0084] In an embodiment, the object detection model includes an encoder, a memory module, a feature fusion module, a spatial attention module, and a decoder that are sequentially connected. The step of inputting the original image information into the object detection model to output the object position information of the original image information through the object detection model includes: inputting the original image information into the encoder to extract features to obtain first feature information; inputting the first feature information into the memory module to fuse the first feature information and memory information through the memory module to obtain a plurality of second feature information of different scales; inputting the plurality of second feature information of different scales into the feature fusion module to fuse features to obtain a plurality of fusion feature information; inputting the plurality of fusion feature information into the spatial attention module for processing to obtain a plurality of third feature information; and inputting the plurality of third feature information into the decoder for decoding to obtain the object position information of the original image information.

[0085] Further, the memory module includes a first feature extraction layer, a first up-sampling layer, a first convolution layer, a first fusion layer, a second up-sampling layer, a second convolution layer, a second fusion layer, a second feature extraction layer, and a third feature extraction layer. The first feature extraction layer, the first up-sampling layer, the first convolution layer, the first fusion layer, the second up-sampling layer, the second convolution layer, and the second fusion layer are sequentially cascaded. An output item of the second feature extraction layer is an input item of the first fusion layer, and an output item of the third feature extraction layer is an input item of the second fusion layer. The first convolution layer and the second convolution layer can use a 3x3 convolution kernel. The multiple second feature information of different scales includes a first sub-feature, a second sub-feature, and a third sub-feature. The multiple fusion feature information includes a first fusion feature information, a second fusion feature information, and a third fusion feature information. The multiple second feature information of different scales is input into the feature fusion module for feature fusion to obtain the multiple fusion feature information, including: inputting the first sub-feature into the first feature extraction layer for feature extraction to obtain the first fusion feature information; inputting the first fusion feature information into the first up-sampling layer for up-sampling processing to obtain a first up-sampling feature; inputting the first up-sampling feature into the first convolution layer for convolution processing to obtain a first convolution feature; inputting the second sub-feature into the second feature extraction layer for feature extraction to obtain a fourth sub-feature; inputting the fourth sub-feature and the first convolution feature into the first fusion layer for fusion to obtain the second fusion feature information; inputting the second fusion feature information into the second up-sampling layer for up-sampling processing to obtain a second up-sampling feature; inputting the second up-sampling feature into the second convolution layer for convolution processing to obtain a second convolution feature; inputting the third sub-feature into the third feature extraction layer for feature extraction to obtain a fifth sub-feature; and inputting the fifth sub-feature and the second convolution feature into the second fusion layer for fusion to obtain the third fusion feature information.

[0086] The first feature extraction layer includes a third convolution layer, a third fusion layer, a fourth convolution layer, and a first attention layer connected in parallel with the third convolution layer. The step of inputting the first sub-feature into the first feature extraction layer for feature extraction to obtain the first fusion feature information includes: inputting the first sub-feature into the third convolution layer for convolution processing to obtain a third convolution feature; inputting the first sub-feature into the first attention layer for processing to obtain first channel information; inputting the third convolution feature and the first channel information into the third fusion layer for multiplication processing to obtain a first fusion feature; and inputting the first fusion feature into the fourth convolution layer for convolution processing to obtain the first fusion feature information.

[0087] The second feature extraction layer includes a fifth convolutional layer, a fourth fusion layer, a sixth convolutional layer and a second attention layer connected in parallel with the fifth convolutional layer, and the step of inputting the second sub-feature into the second feature extraction layer for feature extraction to obtain the fourth sub-feature includes: inputting the second sub-feature into the fifth convolutional layer for convolution processing to obtain a fourth convolutional feature; inputting the second sub-feature into the second attention layer for processing to obtain second channel information; inputting the fourth convolutional feature and the second channel information into the fourth fusion layer for multiplication processing to obtain a second fusion feature; and inputting the second fusion feature into the sixth convolutional layer for convolution processing to obtain the fourth sub-feature.

[0088] The third feature extraction layer includes a seventh convolutional layer, a fifth fusion layer, an eighth convolutional layer and a third attention layer connected in parallel with the seventh convolutional layer, and the step of inputting the third sub-feature into the third feature extraction layer for feature extraction to obtain the fifth sub-feature includes: inputting the third sub-feature into the seventh convolutional layer for convolution processing to obtain a fifth convolutional feature; inputting the third sub-feature into the third attention layer for processing to obtain third channel information; inputting the fifth convolutional feature and the third channel information into the fifth fusion layer for multiplication processing to obtain a third fusion feature; and inputting the third fusion feature into the eighth convolutional layer for convolution processing to obtain the fifth sub-feature.

[0089] The spatial attention module includes a first pooling layer, a third up-sampling layer, a sixth fusion layer, a fourth up-sampling layer and a seventh fusion layer connected in sequence, and the plurality of third feature information includes a sixth sub-feature, a seventh sub-feature and an eighth sub-feature, and the step of inputting the plurality of fusion feature information into the spatial attention module for processing to obtain the plurality of third feature information includes: inputting the first fusion feature information into the first pooling layer for pooling processing to obtain the sixth sub-feature; inputting the sixth sub-feature into the third up-sampling layer for up-sampling processing to obtain a third up-sampling feature; inputting the third up-sampling feature and the second fusion feature information into the sixth fusion layer for multiplication processing to obtain the seventh sub-feature; inputting the seventh sub-feature into the fourth up-sampling layer for up-sampling processing to obtain a fourth up-sampling feature; and inputting the fourth up-sampling feature and the seventh sub-feature into the seventh fusion layer for multiplication processing to obtain the eighth sub-feature.

[0090] In an embodiment, before the step of inputting the original image information into the object detection model in the step S401 and outputting the object position information of the original image information by the object detection model, the method further includes: obtaining a training image and a memory sample; generating a simulation image based on the training image; performing feature extraction on the memory sample to obtain memory information; and training a preset network model based on the training image, the simulation image and the memory information to obtain the object detection model. The structure of the preset network model is the same as the model structure of the object detection model, and the model structure of the object detection model can be referred to the foregoing description.

[0091] In one specific embodiment, the step of generating a simulated image based on a training image specifically includes: acquiring two-dimensional Palin noise and binarizing it to obtain a noise mask; binarizing the training image to obtain a mask image; multiplying the mask image and the noise mask to obtain a first processed image; determining a noise foreground image based on the first processed image and a pre-acquired noise image; inverting the first processed image and multiplying the inverted first processed image with the training image to obtain a second processed image; and superimposing the noise foreground image and the second processed image to obtain a simulated image. The process of determining the noise foreground image based on the first processed image and the pre-acquired noise image can be represented as: I′ n =δ(M⊙I n )+(1-δ)(M⊙I),I′ n M represents the noisy foreground image, and I represents the first processed image. n δ represents the noisy image, I represents the sample image, and δ represents the preset transparency coefficient.

[0092] S402. Determine the target image detection result based on object location information and region location information.

[0093] The object location information refers to the location of the object in the original image information, while the region location information refers to the location of the intersecting region of the intersecting components. Therefore, based on the object location information and the location information of the intersecting region, it can be determined whether an object exists in the intersecting region. For example, if the object location information is the location information of a defect in the original image information, it can be determined whether a defect exists in the intersecting region based on the location information of the defect and the location information of the intersecting region.

[0094] In one specific embodiment, the step of determining the target image detection result based on object location information and region location information specifically includes: determining whether there is overlapping location information between the object location information and the region location information; if there is overlapping location information between the object location information and the region location information, determining that an object exists in the intersecting region; if there is no overlapping location information between the object location information and the region location information, determining that no object exists in the intersecting region. For example, if there is overlapping location information (x1, y1) between the object location information and the region location information, then it is determined that there is a defect in the intersecting region.

[0095] It should be noted that the object location information may include only one location information or multiple location information. When the object location information includes multiple location information, if at least one location information of the object location information overlaps with the area location information, it can be considered that the object location information and the area location information have overlapping location information.

[0096] In a specific embodiment, the target image detection result is an image detection result of the panel to be detected, and after the target image detection result is determined based on the original image information and the region position information in step S203, the method comprises: if the target image detection result includes an object in the intersection region, converting the overlapping position information of the object position information and the region position information to obtain the to-be-repaired position information of the panel to be detected; and performing repair processing on the panel to be detected based on the to-be-repaired position information.

[0097] In the embodiments of the present application, the overlapping position information of the object position information and the region position information is defect position information of the component intersection region of the panel to be detected in the original image information, and the overlapping position information is converted based on the module parameters of the imaging module, so that the defect position information is converted from the image plane to the spatial plane, thereby obtaining the position information of the panel to be detected that needs to be repaired, and the repair processing is performed on the panel to be detected based on the to-be-repaired position information, thereby improving the defect repair efficiency of the panel to be detected.

[0098] In summary, the object detection method provided by the embodiments of the present application comprises: obtaining original image information, performing region positioning on the original image information to obtain region position information, and determining a target image detection result based on the original image information and the region position information. In the present scheme, the position information of the component intersection region in the original image information is obtained by performing region positioning on the original image information, and then it is determined whether there is an object in the intersection region based on the original image information and the region position information, so that accurate defect detection can be achieved for display panels of different resolutions and sizes. The object detection method has strong universality and expandability. Furthermore, the first filtering processing is performed on the original image information to obtain first filtered image information, the second filtering processing is performed on the first filtered image information to obtain second filtered image information, the region position information is determined based on the second filtered image information, which can adapt to key component positioning in a complex scene and improve the accuracy of the defect detection result. The MemSeg model is used for defect detection on the original image information, which can solve the problems of sample imbalance and difficulty in collecting defect data in a complex scene, thereby improving the precision and robustness of the defect detection result.

[0099] In order to better implement the image detection method in the embodiments of the present application, based on the image detection method, the embodiments of the present application further provide an image detection device, as shown in Figure 8 The image detection device 600 comprises:

[0100] An image acquisition module 610 is configured to acquire original image information.

[0101] A region positioning module 620 is configured to perform region positioning on the original image information to obtain region position information.

[0102] The image detection module 630 is configured to determine a target image detection result based on the original image information and the region position information.

[0103] In the embodiments of the present application, the position information of the intersection region of the components in the original image information is obtained by region positioning of the original image information, and then whether the intersection region has an object is determined based on the original image information and the region position information, so that accurate defect detection can be achieved for display panels with different resolutions and sizes.

[0104] In some embodiments of the present application, the original image information includes a plurality of first candidate regions and a plurality of second candidate regions, the plurality of first candidate regions are regions extending in the horizontal direction, and the plurality of second candidate regions are regions extending in the vertical direction. The region positioning module 620 performs region positioning on the original image information to obtain the region position information, including:

[0105] determining a first target region from the plurality of first candidate regions based on first gradient information corresponding to the first candidate region; the first target region is a first candidate region with first gradient information greater than a preset first threshold value;

[0106] determining a second target region from the plurality of second candidate regions based on second gradient information corresponding to the second candidate region; the second target region is a second candidate region with second gradient information greater than a preset second threshold value;

[0107] determining the region position information based on the first target region and the second target region.

[0108] In some embodiments of the present application, before the region positioning module 620 performs region positioning on the original image information to obtain the region position information, the region positioning module 620 updates the original image information based on the following manner:

[0109] performing first filtering processing on the original image information to obtain first filtered image information;

[0110] performing second filtering processing on the first filtered image information to obtain new original image information.

[0111] In some embodiments of the present application, the region positioning module 620 performs second filtering processing on the first filtered image information to obtain new original image information, including:

[0112] performing threshold segmentation on the first filtered image information to obtain region contour information;

[0113] performing connected domain detection on the region contour information to obtain connected domain information;

[0114] filtering the region contour information based on the connected domain information to obtain new original image information.

[0115] In some embodiments of the present application, the region positioning module 620 determines the region position information based on the first target region and the second target region, including

[0116] determining the first position information based on the first target region; the first position information is position information corresponding to an edge point of the first target region;

[0117] determining the second position information based on the second target region; the second position information is position information corresponding to an edge point of the second target region;

[0118] determining the region position information based on the first position information and the second position information.

[0119] In some embodiments of the present application, the first gradient information includes horizontal gradient information and / or vertical gradient information, and the first gradient information of the first candidate region is obtained by the region positioning module performing the following steps:

[0120] For any first candidate region in the plurality of first candidate regions, calculating a pixel difference value of horizontally adjacent pixel points in the first candidate region;

[0121] summing the pixel difference values of the horizontally adjacent pixel points to obtain a first pixel accumulation value;

[0122] dividing the first pixel accumulation value by the number of pixels in the first candidate region to obtain the horizontal gradient information of the first candidate region; and / or,

[0123] For any first candidate region in the plurality of first candidate regions, calculating a pixel difference value of vertically adjacent pixel points in the first candidate region;

[0124] summing the pixel difference values of the vertically adjacent pixel points to obtain a second pixel accumulation value;

[0125] dividing the second pixel accumulation value by the number of pixels in the first candidate region to obtain the vertical gradient information of the first candidate region.

[0126] In some embodiments of the present application, the image detection module 630 determines the target image detection result based on the original image information and the region position information, including:

[0127] inputting the original image information into the object detection model to output object position information of the original image information by the object detection model;

[0128] determining the target image detection result based on the object position information and the region position information.

[0129] In some embodiments of the present application, the target image detection result includes an object existing in the intersection region and an object not existing in the intersection region, and the image detection module 630 determines the target image detection result based on the object position information and the region position information, including:

[0130] determining whether the object position information and the region position information have overlapping position information;

[0131] if the object position information and the region position information have overlapping position information, determining that the object exists in the intersection region;

[0132] if the defect position information and the region position information do not have overlapping position information, determining that the object does not exist in the intersection region.

[0133] In some embodiments of the present application, the target image detection result is the image detection result of the panel to be detected, and after the image detection module 630 determines the target image detection result based on the original image information and the region position information, the image detection module 630 is further configured to:

[0134] if the target image detection result includes an object existing in the intersection region, converting the overlapping position information of the object position information and the region position information to obtain the repair position information of the panel to be detected;

[0135] performing repair processing on the panel to be detected based on the repair position information.

[0136] The embodiments of the present application also provide a computer device integrating any one of the object detection apparatuses provided by the embodiments of the present application. The computer device includes:

[0137] one or more processors;

[0138] a memory; and

[0139] one or more application programs, wherein the one or more application programs are stored in the memory and are configured to perform the steps of the object detection method in any one of the image detection method embodiments by the processor.

[0140] The embodiments of the present application also provide a computer device integrating any one of the image detection apparatuses provided by the embodiments of the present application. As shown in FIG. 8, it shows a structural schematic diagram of the computer device related to the embodiments of the present application, specifically: Figure 9

[0141] The computer device can include a processor 801 with one or more processing cores, a memory 802 with one or more computer readable storage media, a power supply 803, and an input unit 804, and the like. Those skilled in the art can understand that, Figure 9 ​The computer device structure shown in the figure is not a limitation of the computer device, and can include more or fewer components than shown, or combine certain components, or arrange different components. Among them:

[0142] The processor 801 is the control center of the computer device, connects various parts of the computer device through various interfaces and lines, and performs various functions of the computer device and processes data by running or executing software programs and / or modules stored in the memory 802 and calling data stored in the memory 802, thereby overall monitoring the computer device. Optionally, the processor 801 can include one or more processing cores; preferably, the processor 801 can integrate an application processor and a modem processor, wherein the application processor mainly processes the operating system, user interface, and application program, etc., and the modem processor mainly processes wireless communication. It can be understood that the above-mentioned modem processor can also not be integrated into the processor 801.

[0143] The memory 802 can be used to store software programs and modules, and the processor 801 executes various functions and data processing by running the software programs and modules stored in the memory 802. The memory 802 can mainly include a program storage area and a data storage area, wherein the program storage area can store an operating system, at least one application program required by a function (such as a sound playing function, an image playing function, etc.), etc.; the data storage area can store data created according to the use of the computer device, etc. In addition, the memory 802 can include a high-speed random access memory, and can also include a non-volatile memory, such as at least one magnetic disk storage device, a flash memory device, or other volatile solid-state memory device. Accordingly, the memory 802 can also include a memory controller to provide access for the processor 801 to the memory 802.

[0144] The computer device also includes a power supply 803 for powering various components, and preferably the power supply 803 can be logically connected to the processor 801 through a power management system, so as to realize functions such as management of charging, discharging, and power consumption management through the power management system. The power supply 803 can also include one or more direct current or alternating current power supplies, a recharging system, a power failure detection circuit, a power converter or inverter, a power state indicator, and any other components.

[0145] The computer device can also include an input unit 804, which can be used to receive input digital or character information, and generate keyboard, mouse, joystick, optical or trackball signal inputs related to user settings and function control.

[0146] Although not shown, the computer device can further include a display unit and the like, which will not be described here. Specifically in the embodiment, the processor 801 in the computer device will load the executable file corresponding to the process of one or more application programs into the memory 802 according to the following instructions, and run the application program stored in the memory 802 by the processor 801, thereby realizing various functions, as follows:

[0147] obtaining original image information;

[0148] performing region positioning on the original image information to obtain region position information;

[0149] determining a target image detection result based on the original image information and the region position information.

[0150] Those skilled in the art can understand that all or part of the steps in the various methods of the above embodiments can be completed by instructions, or by related hardware controlled by the instructions, which can be stored in a computer readable storage medium and loaded and executed by a processor.

[0151] To this end, the embodiments of the present application provide a computer readable storage medium, which can include a read only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk, etc. A computer program is stored on the storage medium, and the computer program is loaded by a processor to execute the steps in any of the image detection methods provided by the embodiments of the present application. For example, the computer program loaded by the processor can execute the following steps:

[0152] obtaining original image information;

[0153] performing region positioning on the original image information to obtain region position information;

[0154] determining a target image detection result based on the original image information and the region position information.

[0155] In the above embodiments, the description of each embodiment has its own focus, and the parts not described in detail in a certain embodiment can be referred to the detailed description of other embodiments above, which will not be described here.

[0156] In specific implementation, each of the above units or structures can be implemented as an independent entity, or can be combined as the same or several entities, and the specific implementation of each of the above units or structures can be referred to the method embodiments above, which will not be described here.

[0157] The specific implementation of each of the above operations can be referred to the embodiments above, which will not be described here.

[0158] The above describes in detail the image detection method, device, computer device and computer readable storage medium provided by the embodiments of the present application. The principles and implementation manners of the present application are described by applying specific examples. The above description of the embodiments is only used to help understand the method of the present application and its core idea. Meanwhile, for those skilled in the art, according to the idea of the present application, the specific implementation manners and application ranges will be changed. In conclusion, the content of the specification should not be understood as a limitation of the present application.

Claims

1. An image detection method, characterized in that, include: Obtain the original image information; The original image information is used to perform region localization to obtain region location information; Based on the original image information and the region location information, the target image detection result is determined.

2. The method according to claim 1, characterized in that, The original image information includes a number of first candidate regions and a number of second candidate regions, wherein the number of first candidate regions are regions extending horizontally and the number of second candidate regions are regions extending vertically. The step of locating a region from the original image information to obtain region location information includes: A first target region is determined from a plurality of first candidate regions based on the first gradient information corresponding to the first candidate region; the first target region is a first candidate region whose first gradient information is greater than a preset first threshold. A second target region is determined from several second candidate regions based on the second gradient information corresponding to the second candidate region; the second target region is a second candidate region whose second gradient information is greater than a preset second threshold. Based on the first target area and the second target area, determine the area location information.

3. The method according to claim 2, characterized in that, Before performing region localization on the original image information to obtain region location information, the original image information is updated based on the following method: The original image information is subjected to a first filtering process to obtain the first filtered image information; The first filtered image information is subjected to a second filtering process to obtain new original image information.

4. The method according to claim 3, characterized in that, The second filtering process on the first filtered image information to obtain new original image information includes: The first filtered image information is subjected to threshold segmentation to obtain region contour information; Connectivity detection is performed on the region contour information to obtain connected component information; The region contour information is filtered based on the connected component information to obtain new original image information.

5. The method according to claim 2, characterized in that, The step of determining the region location information based on the first target region and the second target region includes... Based on the first target area, determine the first location information; The first location information is the location information corresponding to the edge point of the first target area; Based on the second target region, second location information is determined; the second location information is the location information corresponding to the edge points of the second target region. Based on the first location information and the second location information, the regional location information is determined.

6. The method according to claim 2, characterized in that, The first gradient information includes lateral gradient information and / or vertical gradient information; the first gradient information of the first candidate region is obtained based on the following method: For any one of the first candidate regions, calculate the pixel difference between horizontally adjacent pixels in the first candidate region. The pixel differences between the horizontally adjacent pixels are summed to obtain the first pixel accumulation value; Divide the accumulated value of the first pixel by the number of pixels in the first candidate region to obtain the lateral gradient information of the first candidate region; And / or, For any one of the first candidate regions, calculate the pixel difference between vertically adjacent pixels in the first candidate region. The pixel differences between the vertically adjacent pixels are summed to obtain the second pixel accumulation value; The vertical gradient information of the first candidate region is obtained by dividing the accumulated value of the second pixel by the number of pixels in the first candidate region.

7. The method according to claim 1, characterized in that, The step of determining the target image detection result based on the original image information and the region location information includes: The original image information is input into the object detection model, and the object detection model outputs the object location information of the original image information. Based on the object location information and the region location information, the target image detection result is determined.

8. The method according to claim 7, characterized in that, The target image detection results include the presence of objects in the intersecting regions and the absence of objects in the intersecting regions; The step of determining the target image detection result based on the object location information and the region location information includes: Determine whether there is any overlapping location information between the object location information and the region location information; If the object location information overlaps with the region location information, it is determined that the object exists in the intersecting region. If the object location information and the region location information do not overlap, it is determined that the object does not exist in the intersecting region.

9. The method according to claim 8, characterized in that, The target image detection result is the image detection result of the panel to be detected. After determining the target image detection result based on the original image information and the region location information, the process includes: If the target image detection result includes the presence of an object in the intersecting region, the overlapping position information of the object position information and the region position information is converted to obtain the repair position information of the panel to be detected; The panel to be inspected is repaired based on the location information to be repaired.

10. An image detection device, characterized in that, include: The image acquisition module is used to acquire raw image information; The region positioning module is used to perform region positioning on the original image information to obtain region location information; An image detection module is used to determine the target image detection result based on the original image information and the region location information; Preferably, the original image information includes a plurality of first candidate regions and a plurality of second candidate regions, wherein the plurality of first candidate regions are regions extending horizontally, and the plurality of second candidate regions are regions extending vertically. The region localization module performs region localization on the original image information to obtain region location information, including: A first target region is determined from a plurality of first candidate regions based on the first gradient information corresponding to the first candidate region; the first target region is a first candidate region whose first gradient information is greater than a preset first threshold. A second target region is determined from several second candidate regions based on the second gradient information corresponding to the second candidate region; the second target region is a second candidate region whose second gradient information is greater than a preset second threshold. Based on the first target area and the second target area, determine the area location information; Preferably, before the region positioning module performs region positioning on the original image information to obtain region location information, the region positioning module also updates the original image information based on the following method: The original image information is subjected to a first filtering process to obtain the first filtered image information; The first filtered image information is subjected to a second filtering process to obtain new original image information; Preferably, the region positioning module performs a second filtering process on the first filtered image information to obtain new original image information, including: The first filtered image information is subjected to threshold segmentation to obtain region contour information; Connectivity detection is performed on the region contour information to obtain connected component information; Based on the connected component information, the region contour information is filtered to obtain new original image information; Preferably, the area positioning module determines area location information based on the first target area and the second target area, including: Based on the first target region, first location information is determined; the first location information is the location information corresponding to the edge point of the first target region. Based on the second target region, second location information is determined; the second location information is the location information corresponding to the edge points of the second target region. Based on the first location information and the second location information, determine the area location information; Preferably, the first gradient information includes lateral gradient information and / or vertical gradient information, and the first gradient information of the first candidate region is obtained by the region localization module through the following steps: For any one of the first candidate regions, calculate the pixel difference between horizontally adjacent pixels in the first candidate region. The pixel differences between the horizontally adjacent pixels are summed to obtain the first pixel accumulation value; Divide the accumulated value of the first pixel by the number of pixels in the first candidate region to obtain the lateral gradient information of the first candidate region; and / or, For any one of the first candidate regions, calculate the pixel difference between vertically adjacent pixels in the first candidate region. The pixel differences between the vertically adjacent pixels are summed to obtain the second pixel accumulation value; The vertical gradient information of the first candidate region is obtained by dividing the accumulated value of the second pixel by the number of pixels in the first candidate region. Preferably, the image detection module determines the target image detection result based on the original image information and the region location information, including: The original image information is input into the object detection model, and the object detection model outputs the object location information of the original image information. Based on the object location information and the region location information, the target image detection result is determined; Preferably, the target image detection result includes the presence of an object in the intersecting region and the absence of the object in the intersecting region. The image detection module determines the target image detection result based on the object location information and the region location information, including: Determine whether there is any overlapping location information between the object location information and the region location information; If the object location information overlaps with the region location information, it is determined that the object exists in the intersecting region. If the object location information and the region location information do not overlap, it is determined that the object does not exist in the intersecting region. The target image detection result is the image detection result of the panel to be detected. After the image detection module determines the target image detection result based on the original image information and the region location information, it includes: If the target image detection result includes the presence of an object in the intersecting region, the overlapping position information of the object position information and the region position information is converted to obtain the repair position information of the panel to be detected; The panel to be inspected is repaired based on the location information to be repaired.

11. A computer device, characterized in that, The computer device includes: One or more processors; Memory; and One or more applications, wherein the one or more applications are stored in the memory and configured to be executed by the processor to implement the image detection method of any one of claims 1 to 9.

12. A computer-readable storage medium, characterized in that, It stores a computer program, which is loaded by a processor to perform the steps of the image detection method according to any one of claims 1 to 9.

Images