Panel defect detection method and device, computer device and storage medium

Abstract

The application provides a panel defect detection method and device, computer equipment and a storage medium. The method comprises: acquiring a to-be-detected image of a to-be-detected panel; generating luminosity template images corresponding to different sub-pixel offset values based on a reference light-emitting area and a reference light-emitting area in the to-be-detected image; determining the sub-pixel offset value corresponding to each light-emitting unit according to the light-emitting area in the to-be-detected image and the luminosity template images; and determining a defective light-emitting unit in the to-be-detected panel based on the defect detection result of the to-be-detected image and the sub-pixel offset value corresponding to each light-emitting unit. The application determines the sub-pixel offset value corresponding to each light-emitting unit by using the luminosity template images corresponding to different sub-pixel offset values, can obtain more accurate position information of the light-emitting unit with defects, and thus improves the effect of panel defect detection.

Description

Technical Field

[0001] This application relates to the technical field, specifically to a panel defect detection method, apparatus, computer equipment, and storage medium (computer-readable storage medium). Background Technology

[0002] A display panel consists of several closely arranged light-emitting units. Before leaving the factory, it needs to be inspected for display defects to ensure normal display during use. Typically, defect detection in display panels involves taking pictures of the light-emitting areas on the panel with a camera. Each light-emitting unit on the panel forms a regular bright spot on the image. Therefore, locating a defective light-emitting unit requires first determining the position and size of each light-emitting unit on the image.

[0003] However, in actual testing, there will be sub-pixel mismatch in the light spots formed between adjacent light-emitting units. In addition, considering the distortion of the camera lens and the characteristics of imaging, there will be slight differences in the sub-pixel mismatch of light-emitting units in different areas. This will affect the positioning accuracy of the light-emitting units in the image to a certain extent, thus affecting the subsequent defect detection of the light-emitting units. Summary of the Invention

[0004] Therefore, it is necessary to provide a panel defect detection method, apparatus, computer equipment, and storage medium to address the aforementioned technical problems, thereby resolving the technical issue of insufficient positioning accuracy of light-emitting units in existing panel defect detection methods, which affects the defect detection effect.

[0005] In a first aspect, this application provides a panel defect detection method, including:

[0006] Obtain the image to be tested of the panel to be tested; the image to be tested includes the light-emitting area corresponding to the light-emitting unit on the panel to be tested;

[0007] Based on the reference luminous region in the image to be detected and the reference luminous region corresponding to the reference luminous region, multiple photometric template images corresponding to different sub-pixel offset values ​​are generated.

[0008] Based on the luminous region corresponding to the luminous unit in the image to be detected and the multiple photometric template images, determine the sub-pixel offset value corresponding to each luminous unit;

[0009] Based on the defect detection results of the image to be detected and the sub-pixel offset value corresponding to each of the light-emitting units, the defective light-emitting units in the panel to be detected are determined.

[0010] As a feasible embodiment of this application, the step of generating multiple photometric template images corresponding to different sub-pixel offset values ​​based on the reference luminous region in the image to be detected and the reference luminous region corresponding to the reference luminous region includes:

[0011] Corner detection is performed on the image to be detected to obtain the target corners in the image to be detected;

[0012] The reference luminous area and the corresponding reference luminous area are determined based on the coordinate information of the target corner point.

[0013] Based on the reference luminous region and the reference luminous region, multiple photometric template images corresponding to different sub-pixel offset values ​​are generated.

[0014] As a feasible embodiment of this application, determining the reference luminous region and the corresponding reference luminous region based on the coordinate information of the target corner point includes:

[0015] Obtain the width and height of the region corresponding to the light-emitting unit in the image to be detected;

[0016] Based on the coordinate information of the target corner point and the width and height of the region, the reference luminous area is determined;

[0017] The reference luminescent region is obtained by translating the reference luminescent region according to the width and height of the region.

[0018] As a feasible embodiment of this application, the step of generating multiple photometric template images corresponding to different sub-pixel offset values ​​based on the reference luminous region and the reference luminous region includes:

[0019] An initial photometric image is generated based on the pixel value difference between corresponding pixels in the reference luminous region and the reference luminous region.

[0020] The initial photometric image is interpolated based on a preset subpixel offset value to obtain multiple photometric template images corresponding to different subpixel offset values.

[0021] As a feasible embodiment of this application, the reference light-emitting region includes a first reference light-emitting region, a second reference light-emitting region, a third reference light-emitting region, and a fourth reference light-emitting region;

[0022] The step of generating an initial luminous image based on the pixel value difference between corresponding pixels in the reference luminous region and the reference luminous region includes:

[0023] The pixel value difference between the reference luminescent region and the corresponding pixels of the first, second, third, and fourth reference luminescent regions is calculated respectively to obtain the first photometric image, the second photometric image, the third photometric image, and the fourth photometric image.

[0024] An initial photometric image is generated based on the pixel value relationship of corresponding pixels in the first, second, third, and fourth photometric images.

[0025] As a feasible embodiment of this application, determining the sub-pixel offset value corresponding to each light-emitting unit based on the light-emitting region corresponding to the light-emitting unit in the image to be detected and the plurality of photometric template images includes:

[0026] Calculate the correlation coefficient between the target luminescent region in the image to be detected and each of the photometric template images;

[0027] Based on the magnitude relationship between the correlation coefficients, the target template image is determined from the photometric template image;

[0028] The subpixel offset value corresponding to the target template image is determined as the subpixel offset value of the light-emitting unit corresponding to the target light-emitting region.

[0029] As a feasible embodiment of this application, the target light-emitting area includes light-emitting areas corresponding to multiple target light-emitting units;

[0030] The step of determining the sub-pixel offset value corresponding to the target template image as the sub-pixel offset value of the light-emitting unit corresponding to the target light-emitting region includes:

[0031] The sub-pixel offset value corresponding to the target template image is determined as the sub-pixel offset value corresponding to the target luminous region;

[0032] Based on the light-emitting unit corresponding to the target light-emitting region, the sub-pixel offset value corresponding to the target light-emitting region is interpolated to obtain the sub-pixel offset value corresponding to each target light-emitting unit.

[0033] As a feasible embodiment of this application, the defect detection results include point defect detection results and / or Mura defect detection results;

[0034] Before determining the defective light-emitting units in the panel to be detected based on the defect detection results of the image to be detected and the sub-pixel offset values ​​corresponding to each of the light-emitting units, the method further includes:

[0035] Obtain the image to be detected and the corresponding translational image of the image to be detected;

[0036] Based on the pixel value difference between the corresponding pixels in the translated image and the image to be detected, a residual image of luminance distribution is obtained;

[0037] Based on the average and variance of the pixel values ​​of each pixel in the residual image of the photometric distribution, abnormal pixels are extracted from the residual image of the photometric distribution.

[0038] Based on the abnormal pixels, the point defect detection result is determined; and / or,

[0039] The image to be detected is processed based on a preset point spread function to obtain a photometric distribution image corresponding to the image to be detected;

[0040] The photometric distribution image is binarized based on the average and variance of the pixel values ​​of each preset region in the photometric distribution image to obtain a binarized image.

[0041] The Mura defect detection result of the image to be detected is determined based on the connected regions in the binarized image.

[0042] Secondly, this application provides a panel defect detection device, comprising:

[0043] An acquisition module is used to acquire an image of the panel to be inspected; the image to be inspected includes the light-emitting area corresponding to the light-emitting unit on the panel to be inspected;

[0044] The generation module is used to generate multiple photometric template images corresponding to different sub-pixel offset values ​​based on the reference luminous region in the image to be detected and the reference luminous region corresponding to the reference luminous region.

[0045] The determining module is used to determine the sub-pixel offset value corresponding to each light-emitting unit based on the light-emitting area corresponding to the light-emitting unit in the image to be detected and the plurality of photometric template images;

[0046] The detection module is used to determine the defective light-emitting units in the panel to be detected based on the defect detection results of the image to be detected and the sub-pixel offset values ​​corresponding to each of the light-emitting units.

[0047] Thirdly, this application also provides a computer device, characterized in that the computer device comprises:

[0048] One or more processors;

[0049] Memory; and

[0050] 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 panel defect detection method provided by any of the above.

[0051] Fourthly, this application also provides a computer-readable storage medium having a computer program stored thereon, the computer program being loaded by a processor to perform the steps of the panel defect detection method provided in any of the above claims.

[0052] Fifthly, embodiments of this application provide a computer program product or computer program that includes computer instructions stored in a computer-readable storage medium. A processor of a computer device reads the computer instructions from the computer-readable storage medium and executes the computer instructions, causing the computer device to perform the panel defect detection method provided in any of the preceding claims.

[0053] The panel defect detection method provided in this application, after acquiring the image of the panel to be detected, generates photometric template images matching different sub-pixel offset values ​​based on the reference luminous region and the reference luminous region in the image to be detected. This facilitates the subsequent determination of the sub-pixel offset value corresponding to each luminous unit by utilizing the matching relationship between the luminous region corresponding to the luminous unit in the image to be detected and each photometric template image. In the subsequent defect detection process of the image to be detected, based on the determined detection results and the sub-pixel offset values ​​corresponding to each luminous unit, more accurate location information of the luminous unit with defects can be obtained, thereby improving the effect of panel defect detection. Attached Figure Description

[0054] To more clearly illustrate the technical solutions in the embodiments of this application, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0055] Figure 1(a) is a schematic diagram of a multi-camera acquisition panel provided in an embodiment of this application;

[0056] Figure 1(b) is a schematic diagram of an image captured by a camera according to an embodiment of this application;

[0057] Figure 2 This is a schematic flowchart illustrating the steps of a panel defect detection method provided in an embodiment of this application.

[0058] Figure 3 A flowchart illustrating the steps for determining a reference luminescent region based on corner points, provided in an embodiment of this application;

[0059] Figure 4A flowchart illustrating the steps for determining a reference luminous region and a standard luminous region based on the width and height of the region corresponding to the luminous unit, as provided in this application embodiment;

[0060] Figure 5 This is a schematic flowchart illustrating the steps for generating a photometric template image, as provided in an embodiment of this application.

[0061] Figure 6 A schematic flowchart illustrating the steps for generating a photometric image based on multiple reference luminescent regions, provided in this application embodiment;

[0062] Figure 7 This is a flowchart illustrating the steps for determining a sub-pixel offset value, as provided in an embodiment of this application.

[0063] Figure 8 This application provides a schematic flowchart illustrating the steps involved in detecting point defects.

[0064] Figure 9 This application provides a schematic flowchart of a Mura defect detection procedure.

[0065] Figure 10 This is a schematic diagram of the structure of a panel defect detection device provided in an embodiment of this application;

[0066] Figure 11 This is a schematic diagram of the structure of a computer device provided in an embodiment of this application. Detailed Implementation

[0067] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0068] In the description of this application, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of indicated technical features. Thus, a feature defined as "first" or "second" may explicitly or implicitly include one or more of the stated features. In the description of this application, "multiple" means two or more, unless otherwise explicitly specified.

[0069] In the description of this application, the word "for example" is used to mean "used as an example, illustration, or description." Any embodiment described as "for example" in this application is not necessarily to be construed as being more preferred or advantageous than other embodiments. The following description is provided to enable any person skilled in the art to make and use the invention. Details are set forth in the following description for purposes of explanation. It should be understood that those skilled in the art will recognize that the invention can be made without using these specific details. In other instances, well-known structures and processes will not be described in detail to avoid obscuring the description of the invention with unnecessary detail. Therefore, the invention is not intended to be limited to the embodiments shown, but is consistent with the broadest scope of the principles and features disclosed in this application.

[0070] This application provides a panel defect detection method, apparatus, computer device, and storage medium. The panel defect detection method is installed as a computer program in the panel defect detection apparatus, which is configured as a processor in the computer device. The panel defect detection apparatus in the computer device executes the panel defect detection method provided in this application by running the computer program corresponding to the panel defect detection method. Specifically, to facilitate understanding of the complete implementation scheme of the panel defect detection method provided in this application, the implementation scenario of the panel defect detection method is first described in detail below.

[0071] To detect defects in the panel's light-emitting units, a high-resolution camera is needed to acquire high-resolution images of the panel. Specifically, during image capture, the camera position and focal length need to be adjusted to ensure that each light-emitting unit's image contains multiple pixels, facilitating subsequent defect detection. For example, it is common practice to ensure that each light-emitting unit's image is larger than 5×5 pixels.

[0072] Based on the above, considering that a single high-resolution camera can only capture a limited number of light-emitting units, and that some display panels have large display areas, making it impossible to capture the entire display panel with a single high-resolution camera, multiple high-resolution cameras can be used to capture the light-emitting units in different areas of the display panel. For example, Figure 1(a) provides a schematic diagram of multi-camera panel image acquisition, in which four cameras capture high-resolution images of the upper left, upper right, lower left, and lower right areas of the panel respectively. Specifically, as shown in Figure 1(b), Figure 1(b) is a schematic diagram of an image captured by a camera according to an embodiment of this application. It can be seen that this schematic diagram is the image captured by camera 1 in Figure 1(a), which is the camera capturing the upper left area of ​​the panel. The left and upper halves of the image are black areas outside the panel, while the lower right area is the light-emitting area corresponding to the light-emitting units on the panel. Each light-emitting unit on the panel exists in the image as multiple pixels. Of course, Figure 1(a) above is only one feasible embodiment. In fact, the number of cameras used to capture the light-emitting units of the panel can be determined based on actual needs, that is, based on the actual panel size and camera performance parameters. It is only necessary to ensure the imaging size of each light-emitting unit in the image. This application embodiment will not be described in detail here.

[0073] After capturing panel images via the camera, the camera transmits the acquired panel images to the panel defect detection device to execute the panel defect detection method provided in this application embodiment, thereby identifying defective light-emitting units in various areas of the panel. Specifically, panel defect detection mainly involves local defect detection and overall defect detection, that is, detecting single or multiple light-emitting units with defects such as weak light-emitting ability or no light-emitting at all, and detecting mura-type defects with uneven light-emitting brightness overall. Further details in this application embodiment will not be elaborated upon here.

[0074] In the implementation scenario of the panel defect detection method provided above, the specific steps of the panel defect detection method provided in this proposal will be described below, specifically as follows: Figure 2 As shown, Figure 2 This application provides a schematic flowchart of a panel defect detection method, specifically including steps S210 to S240:

[0075] S210, acquire the image of the panel to be inspected.

[0076] In this embodiment of the application, as shown in the aforementioned schematic diagram of the implementation scenario, the image to be detected includes the light-emitting areas corresponding to each light-emitting unit on the panel to be detected. These areas are typically captured by a camera, as illustrated in Figure 1(b). However, due to factors such as light scattering and camera lens distortion, the light-emitting areas formed by each light-emitting unit in the image to be detected will have a certain degree of sub-pixel-level deviation from the actual situation.

[0077] S220, based on the reference luminous region in the image to be detected and the reference luminous region corresponding to the reference luminous region, generate multiple photometric template images corresponding to different sub-pixel offset values.

[0078] In this embodiment, the reference luminous region in the image to be detected typically refers to the luminous region corresponding to a luminous unit considered as a reference, while the reference luminous region corresponding to the reference luminous region is typically the luminous region corresponding to adjacent luminous units surrounding the reference luminous unit. Specifically, as a feasible embodiment of this application, in the actual detection process, to facilitate calculation and improve detection efficiency, the reference luminous region and the reference luminous region can be obtained based on the corner points of the panel. Specifically, as... Figure 3 As shown, Figure 3 This application provides a flowchart illustrating the steps for determining a reference luminescent region based on corner points, specifically including steps S310 to S330:

[0079] S310, perform corner detection on the image to be detected to obtain the target corners in the image to be detected.

[0080] In this embodiment, considering the special characteristics of corner points, for example, as shown in Figure 1(b) of the aforementioned implementation scenario, the image is an image captured by a camera capturing the upper left area of ​​the panel, which includes the upper left corner point. The left and upper halves of the corner point are both black areas. Therefore, based on this feature, corner point detection can be performed using a preset corner point filtering kernel to determine the target corner point in the image to be detected. Specifically, the following shows a corner point filtering kernel used for the upper left area image:

[0081] -1 1 1 -1 1 1

[0082] The image to be detected is filtered by the above corner filtering kernel. The point with the largest value in the filtered image is the target corner point, which is the reference luminous point used to obtain the reference luminous area. The coordinates C of this corner point are recorded.

[0083] Furthermore, for images in other regions, corner detection can also be performed in a similar manner, that is, by using different preset corner filtering kernels. This will not be elaborated further in the embodiments of this application.

[0084] S320, determine the reference luminous area and the corresponding reference luminous area based on the coordinate information of the target corner point.

[0085] In this embodiment of the application, after obtaining the coordinate information C of the target corner point, the pixels within a certain range are delineated using this coordinate information C as the center point, which serves as the reference luminous region. Based on this, a certain degree of displacement is made to the coordinate information C to obtain a new reference center point, and pixels within the same range are delineated to obtain the reference luminous region. Specifically, to ensure the effect of the subsequently obtained photometric template image, as a feasible embodiment of this application, the size of the selected reference luminous region and the reference luminous region can be set based on the width and height of the area occupied by a luminous unit in the image. Specifically, for example... Figure 4 As shown, Figure 4 This application provides a flowchart illustrating the steps for determining a reference emitting region and a standard emitting region based on the width and height of the region corresponding to the emitting unit. Specifically, it includes steps S410 to S430:

[0086] S410, obtain the width and height of the region corresponding to the light-emitting unit in the image to be detected.

[0087] In this embodiment, the width and height of the region corresponding to the light-emitting unit refers to the number of pixels contained in the light-emitting region formed by the light-emitting unit in the image to be detected in the horizontal and vertical directions. For example, if the aforementioned light-emitting unit is imaged as N×N pixels in the image, then the width and height of the region corresponding to the light-emitting unit is N.

[0088] S420, Based on the coordinate information of the target corner point and the width and height of the region, determine the reference luminous area.

[0089] In this embodiment of the application, based on the coordinate information C of the target corner point and the width and height N of the region, several pixels within the width and height N of the region can be selected with the coordinate information C as the center to obtain the reference luminous region. For example, as a feasible implementation scheme, the region with corner point coordinates C ± 3 * N can be extracted as the reference luminous region CA.

[0090] S430, the reference luminous area is translated according to the width and height of the area to obtain the reference luminous area.

[0091] In this embodiment, based on the foregoing, by translating the reference emitting region by one unit of width and height, the reference emitting region corresponding to the adjacent emitting unit can be obtained. Specifically, as a feasible embodiment of this application, multiple different reference emitting regions can be obtained by translating in multiple different directions. For example, by translating in the four directions of upward, downward, left, and right, four reference emitting regions, CU, CD, CL, and CR, can be obtained respectively.

[0092] S330, based on the reference luminous region and the reference luminous region, generate multiple photometric template images corresponding to different sub-pixel offset values.

[0093] In this embodiment, after obtaining the reference luminous region and the reference luminous region through the aforementioned method, a photometric distribution template corresponding to the corner point can be generated based on the difference between the reference luminous region and the reference luminous region. Based on this, multiple photometric template images corresponding to different sub-pixel offset values ​​can be obtained by linearly interpolating the photometric distribution template in the horizontal and vertical directions. Specifically, as shown... Figure 5 As shown, Figure 5 This application provides a flowchart illustrating the steps for generating a photometric template image, specifically including steps S510 to S520:

[0094] S510, an initial luminous image is generated based on the pixel value difference between the corresponding pixels in the reference luminous region and the reference luminous region.

[0095] In this embodiment, an initial photometric image is obtained by calculating the pixel value difference between corresponding pixels in the reference luminous region and the reference luminous region, and using this pixel value difference as the value of the pixel. However, it should be noted that the reference luminous region may include multiple regions translated in different directions. For example, the aforementioned four reference luminous regions CU, CD, CL, and CR obtained by translation in four directions (up, down, left, and right), meaning the reference luminous region includes a first reference luminous region, a second reference luminous region, a third reference luminous region, and a fourth reference luminous region, are used. In this case, the initial photometric image is obtained by combining the pixel value differences between the reference luminous region and the corresponding pixels in these reference luminous regions. Specifically, as shown... Figure 6 As shown, Figure 6 This application provides a flowchart illustrating the steps for generating a photometric image based on multiple reference luminescent regions, specifically including steps S610 to S620:

[0096] S610, calculate the pixel value difference between the reference luminous region and the corresponding pixel points of the first reference luminous region, the second reference luminous region, the third reference luminous region and the fourth reference luminous region respectively, to obtain the first photometric image, the second photometric image, the third photometric image and the fourth photometric image.

[0097] In this embodiment of the application, the four reference light-emitting regions CU, CD, CL, and CR are respectively used as the first reference light-emitting region, the second reference light-emitting region, the third reference light-emitting region, and the fourth reference light-emitting region for illustration. At this time, the pixel value difference between the reference light-emitting region and the corresponding pixel points of the first reference light-emitting region, the second reference light-emitting region, the third reference light-emitting region, and the fourth reference light-emitting region is calculated respectively to obtain the first photometric image CAU, the second photometric image CAD, the third photometric image CAL, and the fourth photometric image CAR, where CAU = CA - CU, CAD = CA - CD, CAL = CA - CL, and CAR = CA - CR.

[0098] S620, an initial photometric image is generated based on the pixel value relationship of corresponding pixels in the first photometric image, the second photometric image, the third photometric image, and the fourth photometric image.

[0099] In this embodiment of the application, the formula for calculating the initial photometric image D is as follows:

[0100] D=min(max(CAU,CAD),max(CAL,CAR))

[0101] That is, the smaller of the larger value of the corresponding pixel in the CAU image obtained by moving up and the CAD image obtained by moving down, and the larger value of the corresponding pixel in the CAL image obtained by moving left and the CAR image obtained by moving right, is the similarity of the corresponding pixel in the initial photometric image D.

[0102] S520, the initial photometric image is interpolated based on a preset subpixel offset value to obtain multiple photometric template images corresponding to different subpixel offset values.

[0103] In this embodiment, based on the initial photometric image D obtained above, interpolation processing is performed on the initial photometric image D in the horizontal and vertical directions, i.e., the x-axis and y-axis, to obtain multiple photometric template images corresponding to different sub-pixel offset values. Specifically, the number of interpolations affects the accuracy of the sub-pixel offset values. The more interpolations, the more photometric template images are obtained, and the more accurate the obtained sub-pixel offset values ​​are, but the corresponding computational load will also increase. Therefore, appropriate interpolation parameters can be selected based on actual needs. For example, as an optional embodiment of this application, interpolation can be performed from -1 to 1 on the x-axis and y-axis with a step size of 0.1, resulting in a total of 21*21=441 photometric template images under different sub-pixel offsets.

[0104] S230, based on the luminous region corresponding to the luminous unit in the image to be detected and the plurality of photometric template images, determine the sub-pixel offset value corresponding to each luminous unit.

[0105] In this embodiment, after obtaining multiple photometric template images based on the aforementioned scheme, the obtained multiple photometric template images are further matched and compared with the luminous regions corresponding to each luminous unit in the image to be detected, thereby determining the sub-pixel offset value corresponding to each luminous unit. Matching can be achieved based on correlation coefficients, specifically, such as... Figure 7 As shown, Figure 7 This application provides a flowchart illustrating the steps for determining a sub-pixel offset value, specifically including steps S710 to S730:

[0106] S710, calculate the correlation coefficient between the target luminous region in the image to be detected and each of the photometric template images.

[0107] In this embodiment of the application, the target luminous region in the image to be detected can refer to the target luminous region corresponding to each luminous unit. Of course, in order to save computation, improve computational efficiency, and reduce the error caused by the deviation of a single pixel, it is also possible to combine several luminous units to determine the target luminous region. For example, as a feasible implementation scheme of this application, the luminous unit regions corresponding to 10×10, a total of 100 luminous units, can be selected each time to obtain the target luminous region.

[0108] It should be noted that the target luminous area is determined based on pixel coordinates. For example, taking the example that each luminous unit in the image to be detected is 5×5 pixels in size, then for a certain row or column, 0 to 5 is the luminous area corresponding to the first luminous unit, 6 to 10 is the luminous area corresponding to the second luminous unit, and so on.

[0109] In this embodiment, the correlation coefficient between the target luminescent region and each photometric template image can be calculated using the formula for the Pearson correlation coefficient. Specifically, the formula for calculating the Pearson correlation coefficient r is as follows:

[0110]

[0111] The target luminescent region (photometric template image) contains n corresponding pixels, while X i Let be the pixel value of the i-th pixel in the target luminous region. Specifically, the i-th pixel can be the i-th pixel in the pixel sequence obtained after arranging the n pixels in the target luminous region row by row (or column by column), or it can be the i-th pixel in the pixel sequence obtained by other non-repeating arrangement methods. Y represents the average pixel value of each pixel in the target luminous region. i These are the pixel values ​​of each pixel in the photometric template image. This represents the average pixel value of each pixel in the photometric template image.

[0112] S720, Based on the magnitude relationship between the correlation coefficients, determine the target template image from the photometric template image.

[0113] In this embodiment, the correlation coefficient describes the similarity between the target luminescent region and each photometric template image. Therefore, based on the relationship between the correlation coefficients, the target template image most similar to the target luminescent region can be determined from the photometric template images. Specifically, as a feasible implementation, the photometric template image with the highest correlation coefficient can be used as the target template image.

[0114] S730, the sub-pixel offset value corresponding to the target template image is determined as the sub-pixel offset value of the light-emitting unit corresponding to the target light-emitting region.

[0115] In this embodiment of the application, when the target light-emitting area is the light-emitting area corresponding to a single light-emitting unit, since the target template image is most similar to the target light-emitting area, the sub-pixel offset value corresponding to the target template image is the sub-pixel offset value of the light-emitting unit corresponding to the target light-emitting area.

[0116] Of course, as mentioned above, the target luminescent area may correspond to a combination of multiple luminescent units. For example, the target luminescent area may be a combination of 10×10 luminescent units. In this case, after determining the sub-pixel offset value corresponding to the target luminescent area, the sub-pixel offset value corresponding to each target luminescent unit in the target luminescent area can be obtained by interpolation. The embodiments of this application will not be described in detail here.

[0117] S240, based on the defect detection results of the image to be detected and the sub-pixel offset values ​​corresponding to each of the light-emitting units, determine the defective light-emitting units in the panel to be detected.

[0118] In this embodiment, after determining the sub-pixel offset value corresponding to each light-emitting unit, the defective light-emitting unit in the panel to be inspected can be further determined based on the defect detection result of the image to be inspected and the sub-pixel offset value corresponding to each light-emitting unit. Specifically, the defect detection result in the image to be inspected is usually described by coordinate information. In the absence of distortion, considering that the light-emitting units on the panel to be inspected are arranged regularly, that is, the initial coordinate information corresponding to the light-emitting units in the image to be inspected should be arranged in a regular order. However, due to the existence of lens distortion factors, the imaging of the light-emitting units in the image to be inspected has a certain offset. Therefore, by adding the corresponding sub-pixel offset value to the initial position coordinates of each light-emitting unit, the actual position information of each light-emitting unit in the image to be inspected, that is, the actual coordinates, can be obtained. Based on the comparison result of the actual position information and the defect detection result in the image to be inspected, the defective light-emitting unit in the panel to be inspected can be determined.

[0119] Specifically, defect detection results can be point defect detection results or Mura defect detection results. Specific implementation schemes for different defect detection results will be provided later. For example... Figure 8 As shown, Figure 8 This application provides a flowchart illustrating the steps for detecting point defects, specifically including steps S810 to S830:

[0120] S810, acquire the image to be detected and the translation image corresponding to the image to be detected.

[0121] In this embodiment of the application, based on the aforementioned image to be detected, a new translated image can be obtained by translating the image to be detected by several units. For example, the image can be translated to the right by 4 light-emitting units to obtain a translated image.

[0122] S820, based on the pixel value difference between the corresponding pixel points in the translated image and the image to be detected, a residual image of the photometric distribution is obtained.

[0123] In this embodiment, the brightness distribution difference between normal light-emitting units is very small, and only a small amount of luminance remains in the image area after subtraction. However, abnormal light-emitting units differ significantly from normal light-emitting units, resulting in more obvious luminance distribution residue in the image area after subtraction. Therefore, the luminance distribution residue image generated based on the pixel value difference between corresponding pixels in the translated image and the image to be detected can be further used to screen out abnormal light-emitting units.

[0124] S830, based on the average value and variance of the pixel values ​​of each pixel in the residual photometric distribution image, abnormal pixels are extracted from the residual photometric distribution image and used as the point defect detection result of the image to be detected.

[0125] In this embodiment, the three sigma criterion can be used to extract abnormal pixels from the residual image of photometric distribution. Specifically, the mean and variance σ of the pixel values ​​in the local area can be calculated sequentially, and the luminous areas with brightness greater than mean + 3*σ can be searched as abnormal pixels to obtain the point defect detection result of the image to be detected.

[0126] Furthermore, based on the overall Mura defect detection of the image to be detected, specifically, such as... Figure 9 As shown, Figure 9 This application provides a schematic flowchart of a Mura defect detection procedure, specifically including steps S910 to S930:

[0127] S910, the image to be detected is processed based on a preset point spread function to obtain a photometric distribution image corresponding to the image to be detected.

[0128] In this embodiment, considering the light diffusion effect, that is, the luminance distribution of the light-emitting units on the panel in the image to be detected will satisfy the point spread function (PSF), meaning that because two adjacent light-emitting units are very close, their luminance distributions will partially overlap and intersect. Therefore, in order to more accurately measure the true luminance value of each light-emitting unit, a PSF function fitting method can be used to compensate for the luminance in the local area. Specifically, after rounding the coordinates of each luminescent unit to the nearest integer, the maximum value within the corresponding N×N interval in the image to be detected is found. Then, a 3×3 interval centered on the maximum value is taken to calculate the second moment M and the cumulative luminance S of the image. Finally, based on actual measurements, the PSF function is determined to be approximately a Gaussian function with σ = 1.6. Therefore, the proportion of the PSF function integral in the central 3×3 region to the function integral Ra = ∫3 / ∫ can be calculated. ∫3 is the regional integral of the PSF function in the central 3×3 region, and ∫ is the global function integral of the PSF function. Finally, based on the cumulative luminance S and the proportion Ra of the 3×3 interval, the global luminance value L = S / Ra of the current region is calculated, thus obtaining the luminance distribution image.

[0129] Furthermore, it should be noted that, as an optional embodiment of this application, if the panel is captured by multiple cameras, in order to obtain the overall luminance distribution image of the panel, it is necessary to merge the luminance distribution images corresponding to the multiple images to be detected. However, considering that there are certain differences in the exposure time and hardware parameters of different cameras, the brightness of different luminance distribution images has an overall deviation. Therefore, it is also necessary to compensate and calibrate the different luminance distribution images during merging. Specifically, taking the luminance distribution images captured by the cameras in the four corners as an example, the specific implementation scheme includes the following steps: Select the overlapping parts in the four camera shooting areas, and using the brightness of the upper left and upper right images as a reference, calculate the luminance difference D1 of the same light-emitting unit corresponding to each column of overlapping areas in the lower left and lower right images in turn. Then, add D1 to the luminance of each column in the lower left and lower right images and merge them with the upper left and upper right images respectively to obtain the left half image and the right half image of the screen. Then, using the left half image as a reference, calculate the luminance difference D2 of the same light-emitting unit in each row of overlapping areas in the right half image. Then, add D2 to the luminance of each row in the right half image and merge it with the left image to obtain a complete screen luminance value image I.

[0130] Furthermore, to eliminate undesirable distributions caused by equipment malfunctions in the image to be detected, as another optional embodiment of this application, a background distribution image B caused by equipment malfunctions can be pre-constructed, and the undesirable distributions caused by equipment malfunctions in the photometric value image I can be eliminated based on the background distribution image B to obtain an updated photometric value image. The background distribution image B can be a set of photometric value images (I1, I2….I…) of the panel during multiple detection processes. j )

[0131] Updated, of which I j This represents the image of the panel's luminance detected during the j-th detection process. Specifically, the complete implementation scheme includes the following steps:

[0132] A distribution image of equipment defects is constructed using images of preceding materials of the same product. Each material is added to the new background image after inspection, achieving dynamic background updates. The background construction formula is: B. j =B j-1 *momentum+I j *(1-momentum), where B j B is the newly generated background distribution image during the j-th defect detection process. j-1 This refers to the background distribution image generated during the previous defect detection process, specifically the (j-1)th defect detection process. j This is the photometric image of the panel during the j-th defect detection process. Of course, this photometric image I... jThis could be an image preprocessed using brightness boosting, with the mean value increased to 800. Momentum represents the background update impulse. Since the mean value of all newly added images for background calculation is uniformly 800, therefore B... j The mean is also 800;

[0133] Acquire photometric image I j Calculate the mean I of the entire map. m The photometric image is calibrated by mean to obtain the updated photometric image I′=I. j -I m +800, stretches the mean of the photometric value image to 800;

[0134] The calibrated photometric image I′ and the current background distribution image B are obtained. j-1 The difference is used to eliminate uneven image distribution caused by equipment malfunctions. The specific calculation formula is S = I j –B j-1 +800, obtain the brightness distribution image S after removing background unevenness, stretch it to a mean of 800, and then follow the aforementioned steps to... j Added to the background image construction to complete the background distribution image B j-1 Update

[0135] S920, the photometric distribution image is binarized based on the average value and variance of the pixel values ​​of each preset region in the photometric distribution image to obtain a binarized image.

[0136] In this embodiment, after obtaining the photometric distribution image or removing the photometric distribution image unevenly distributed due to equipment malfunction, the K-σClipping method can be used to detect the Mura. Specifically, it includes the following steps:

[0137] The luminance distribution image S i The algorithm is divided into several small local regions, typically 100×100 pixels in size. Within each local region, the mean and standard deviation of all pixel values ​​are calculated. Pixel values ​​that fall outside the mean ± 3σ range are removed. The above method is repeated for the remaining pixel values. This process is iterated multiple times until all remaining pixel values ​​are within the mean ± 3σ range, thus obtaining the final mean and σ values ​​for the current region.

[0138] The distributions of mean and σ values ​​as a function of region D were calculated using the K-σClipping method for all regions. m and D σ ;

[0139] For distribution D m and Dσ Median filtering is used to smooth and eliminate some singular values, and cubic spline interpolation is used to obtain the D-value of the entire image. m and D σ distributed;

[0140] Using the mean and variance distribution plot, i.e. D m and D σ In the photometric distribution image S after removing background unevenness, all pixels falling outside mean±3σ are set to 255, and the remaining pixels are set to 0, to obtain the binarized image S′.

[0141] S930, determine the Mura defect detection result of the image to be detected based on the connected regions in the binarized image.

[0142] In this embodiment, candidate regions for Mura defects are obtained by searching for connected regions in the binarized image S′. Then, noise is filtered out by the area of ​​the connected regions, leaving the true defect regions. The aforementioned method can effectively eliminate the interference of strong mura on the calculation of mean and σ values, so this method also has a very good detection capability for very weak mura defects.

[0143] The panel defect detection method provided in this application, after obtaining the image of the panel to be detected, generates photometric template images matching different sub-pixel offset values ​​based on the reference luminous region and the reference luminous region in the image to be detected. This facilitates the subsequent determination of the sub-pixel offset value corresponding to each luminous unit by utilizing the matching relationship between the luminous region corresponding to the luminous unit in the image to be detected and each photometric template image. In the subsequent defect detection process of the image to be detected, based on the determined detection results and the sub-pixel offset values ​​corresponding to each luminous unit, more accurate location information of the luminous unit with defects can be obtained, thereby improving the effect of panel defect detection.

[0144] To better understand the panel defect detection method provided in the embodiments of this application, the following will refer to Figures 1 to 10 above. Figure 9 The content provides a complete implementation process for a panel defect detection method, which specifically includes the following steps:

[0145] 1) Based on the four preset cameras, high-resolution images of the four panel areas in the upper left, upper right, lower left and lower right are captured respectively to obtain the image to be detected. Ensure that the image of each light-emitting unit in the panel is larger than 5×5 pixels in the image to be detected. Obtain the width and height of the image area occupied by the light-emitting unit, denoted as N.

[0146] 2) The image to be detected is filtered according to the preset corner filter, and the position of the maximum value in the filtered image is determined as the corner coordinate C.

[0147] 3) Extract the region CA with the coordinates of the corner point C ± 3*N. After shifting the corner point C up, down, left, and right by N, extract the four regions CU, CD, CL, and CR in the same way.

[0148] 4) Calculate the differences between CA and the four regions CU, CD, CL, and CR respectively: CAU = CA - CU, CAD = CA - CD, CAL = CA - CL, CAR = CA - CR (where negative numbers are set to 0), and then calculate the photometric image D = min(max(CAU,CAD),max(CAL,CAR)) according to the following formula.

[0149] 5) Using the photometric image D as a template, linear interpolation is performed in the x and y directions to obtain photometric template images under different sub-pixel offsets. The step size of x and y is 0.1 from -1 to 1. A total of 21*21=441 photometric template images under different sub-pixel offsets are obtained.

[0150] 6) Select the region corresponding to 10×10 light-emitting units in a single region matching, and calculate the Pearson correlation coefficient with each photometric template image to obtain the average sub-pixel value of the region corresponding to the 10×10 light-emitting units. Then, obtain the sub-pixel position Ps of each light-emitting unit based on interpolation.

[0151] 7) Point defect detection: After shifting the image to be detected by 4 light-emitting units, the difference between the image and the original image is calculated. The difference in brightness distribution between normal light-emitting units is very small, and there is only a small amount of luminance residue in the image area after the difference. However, abnormal light-emitting points have a large difference from normal light-emitting units, and there will be a more obvious luminance distribution residue in the image area after the difference. The mean and variance σ of the local area are calculated. The light-emitting areas with brightness greater than mean + 3*σ are searched. Combined with the sub-pixel position Ps of each light-emitting unit, the point defect light-emitting units on the panel are determined.

[0152] 8) Mura Defect Detection: Considering the light diffusion effect, the luminance distribution of the light-emitting units on the panel in the image to be detected will satisfy the point spread function (PSF). This means that because two adjacent light-emitting units are very close, their luminance distributions will partially overlap and intersect. Therefore, to more accurately measure the true luminance value of each light-emitting unit, a PSF function fitting method can be used to compensate for the luminance in local areas. Specifically, after rounding the coordinates of each luminescent unit to the nearest integer, the maximum value within the corresponding N×N interval in the image to be detected is found. Then, a 3×3 interval centered on the maximum value is taken to calculate the second moment M and the cumulative luminance S of the image. Finally, based on actual measurements, it is determined that the PSF function approximately satisfies the Gaussian function of σ = 1.6. Therefore, the proportion of the PSF function integral in the central 3×3 region to the function integral Ra = ∫3 / ∫ can be calculated. ∫3 is the regional integral of the PSF function in the central 3×3 region, and ∫ is the global function integral of the PSF function. Finally, based on the cumulative luminance S and the proportion Ra of the 3×3 interval, the global luminance value L = S / Ra of the current region is calculated, thus obtaining the luminance distribution image corresponding to each panel region.

[0153] 9) Select the overlapping areas of the four camera shooting areas. Using the brightness of the upper left and upper right images as a reference, calculate the luminance difference D1 of the same light-emitting unit corresponding to each column of overlapping areas in the lower left and lower right images. Then, add D1 to the luminance of each column in the lower left and lower right images and merge them with the upper left and upper right images respectively to obtain the left and right half images of the screen. Then, using the left half image as a reference, calculate the luminance difference D2 of the same light-emitting unit in each row of overlapping areas in the right half image. Then, add D2 to the luminance of each row in the right half image and merge it with the left image to obtain a complete screen luminance value image I. The luminance value image I obtained in the j-th detection process is denoted as Ij. j

[0154] 10) Construct a distribution image of equipment defects using images of previous materials of the same product. Each material is added to the new background after inspection, achieving dynamic background updates. Background construction formula: B j =B j-1 *momentum+I j *(1-momentum), where B j B is the newly generated background distribution image during the j-th defect detection process. j-1 This refers to the background distribution image generated during the previous defect detection process, specifically the (j-1)th defect detection process. j This is the photometric image of the panel during the j-th defect detection process. Of course, this photometric image I... jThis could be an image preprocessed using brightness boosting, with the mean value increased to 800. Momentum represents the background update impulse. Since the mean value of all newly added images for background calculation is uniformly 800, therefore B... j The mean is also 800;

[0155] 11) Obtain photometric value image I j Calculate the mean I of the entire map. m The photometric image is calibrated by mean to obtain the updated photometric image I′=I. j -I m +800, stretches the mean of the photometric value image to 800;

[0156] 12) Obtain the calibrated photometric image I′ and the current background distribution image B. j-1 The difference is used to eliminate uneven image distribution caused by equipment malfunctions. The specific calculation formula is S=I′–B j-1 +800, obtain the brightness distribution image S after removing background unevenness, stretch it to a mean of 800, and then follow the aforementioned steps to... j Added to the background image construction to complete the background distribution image B j-1 Update;

[0157] 13) Decompose the photometric distribution image S after removing background unevenness into several small local regions, usually 100×100 pixels in size. Calculate the mean and standard deviation of all pixel values ​​in each local region. Remove the pixel values ​​that fall outside mean±3σ. Repeat the above method for the remaining pixel values. Iterate in this way until all the remaining pixel values ​​are within mean±3σ to obtain the final mean and σ values ​​of the current region.

[0158] 14) The distribution D of mean and σ values ​​as a function of region location is calculated using the K-σClipping method for all regions. m and D σ ;

[0159] 15) For distribution D m and D σ Median filtering is used to smooth and eliminate some singular values, and cubic spline interpolation is used to obtain the D-value of the entire image. m and D σ distributed;

[0160] 16) Using the mean and variance distribution plot, i.e., D m and D σIn the photometric distribution image S after removing background unevenness, all pixels falling outside mean±3σ are set to 255, and the remaining pixels are set to 0, to obtain the binarized image S′.

[0161] 17) Find the candidate regions of Mura defects by searching for connected regions in the binarized image S′, and then filter out noise by the area of ​​the connected regions to leave the real defect regions and obtain the Mura defect detection results.

[0162] To better implement the panel defect detection method provided in the embodiments of this application, based on the panel defect detection method provided in the embodiments of this application, this application also provides a panel defect detection device, such as... Figure 10 As shown, the panel defect detection device 1000 includes:

[0163] The acquisition module 1010 is used to acquire the image to be detected of the panel to be detected; the image to be detected includes the light-emitting area corresponding to the light-emitting unit on the panel to be detected;

[0164] The generation module 1020 is used to generate multiple photometric template images corresponding to different sub-pixel offset values ​​based on the reference luminous region in the image to be detected and the reference luminous region corresponding to the reference luminous region.

[0165] The determining module 1030 is used to determine the sub-pixel offset value corresponding to each light-emitting unit based on the light-emitting area corresponding to the light-emitting unit in the image to be detected and the plurality of photometric template images;

[0166] The detection module 1040 is used to determine the defective light-emitting units in the panel to be detected based on the defect detection results of the image to be detected and the sub-pixel offset values ​​corresponding to each of the light-emitting units.

[0167] In some embodiments of this application, the generation module 1020 is further configured to perform corner detection on the image to be detected to obtain target corners in the image to be detected; determine a reference luminous region and a reference luminous region corresponding to the reference luminous region based on the coordinate information of the target corners; and generate multiple photometric template images corresponding to different sub-pixel offset values ​​based on the reference luminous region and the reference luminous region.

[0168] In some embodiments of this application, the generation module 1020 is further configured to obtain the width and height of the region corresponding to the light-emitting unit in the image to be detected; determine the reference light-emitting region based on the coordinate information of the target corner point and the width and height of the region; and translate the reference light-emitting region according to the width and height of the region to obtain the reference light-emitting region.

[0169] In some embodiments of this application, the generation module 1020 is further configured to generate an initial photometric image based on the pixel value difference between corresponding pixels in the reference luminous region and the reference luminous region; and to perform interpolation processing on the initial photometric image based on a preset sub-pixel offset value to obtain multiple photometric template images corresponding to different sub-pixel offset values.

[0170] In some embodiments of this application, the generation module 1020 is further configured to calculate the pixel value difference between the reference luminous region and the corresponding pixels of the first reference luminous region, the second reference luminous region, the third reference luminous region and the fourth reference luminous region respectively, to obtain a first photometric image, a second photometric image, a third photometric image and a fourth photometric image; and generate an initial photometric image based on the size relationship of the pixel values ​​of the corresponding pixels in the first photometric image, the second photometric image, the third photometric image and the fourth photometric image.

[0171] In some embodiments of this application, the determining module 1030 is further configured to calculate the correlation coefficient between the target luminous region in the image to be detected and each of the photometric template images; determine the target template image from the photometric template images based on the magnitude relationship between the correlation coefficients; and determine the sub-pixel offset value corresponding to the target template image as the sub-pixel offset value of the luminous unit corresponding to the target luminous region.

[0172] In some embodiments of this application, the determining module 1030 is further configured to determine the sub-pixel offset value corresponding to the target template image as the sub-pixel offset value corresponding to the target luminous region; and to perform interpolation processing on the sub-pixel offset value corresponding to the target luminous region based on the luminous unit corresponding to the target luminous region to obtain the sub-pixel offset value corresponding to each target luminous unit.

[0173] In some embodiments of this application, before determining the defective light-emitting unit in the panel to be detected based on the defect detection result of the image to be detected and the sub-pixel offset value corresponding to each of the light-emitting units, the detection module 1040 is further configured to acquire the image to be detected and the translation image corresponding to the image to be detected; obtain a photometric distribution residual image based on the pixel value difference between the corresponding pixel points in the translation image and the image to be detected; extract abnormal pixel points from the photometric distribution residual image based on the average value and variance of the pixel values ​​of each pixel point in the photometric distribution residual image; and determine the point defect detection result based on the abnormal pixel points.

[0174] In some embodiments of this application, before determining the defective light-emitting unit in the panel to be detected based on the defect detection result of the image to be detected and the sub-pixel offset value corresponding to each of the light-emitting units, the detection module 1040 is further configured to process the image to be detected based on a preset point diffusion function to obtain a photometric distribution image corresponding to the image to be detected; to perform binarization processing on the photometric distribution image based on the average value and variance of the pixel values ​​of each preset region in the photometric distribution image to obtain a binarized image; and to determine the Mura defect detection result of the image to be detected based on the connected regions in the binarized image.

[0175] Specific limitations regarding the panel defect detection device can be found in the limitations of the panel defect detection method described above, and will not be repeated here. Each module in the aforementioned panel defect detection device can be implemented entirely or partially through software, hardware, or a combination thereof. These modules can be embedded in or independent of the processor in a computer device in hardware form, or stored in the memory of a computer device in software form, so that the processor can call and execute the operations corresponding to each module.

[0176] In some embodiments of this application, the panel defect detection device 1000 can be implemented as a computer program, which can be implemented in, for example... Figure 11 The computer device shown operates on this device. The computer device's memory can store the various program modules that make up the panel defect detection device 1000, for example, Figure 10 The diagram shows an acquisition module 1010, a generation module 1020, a determination module 1030, and a detection module 1040. The computer program comprised of these modules causes the processor to execute the steps in the panel defect detection methods of the various embodiments of this application described in this specification.

[0177] For example, Figure 11 The computer equipment shown can be used as follows Figure 10The acquisition module 1010 in the panel defect detection apparatus 1000 shown executes step S210. The computer device can execute step S220 via the generation module 1020. The computer device can execute step S230 via the determination module 1030. The computer device can execute step S240 via the detection module 1040. The computer device includes a processor, memory, and a network interface connected via a system bus. The processor of the computer device provides computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and internal memory. The non-volatile storage medium stores an operating system and computer programs. The internal memory provides an environment for the operation of the operating system and computer programs in the non-volatile storage medium. The network interface of the computer device is used to communicate with external computer devices via a network connection. When the computer program is executed by the processor, it implements a panel defect detection method.

[0178] Those skilled in the art will understand that Figure 11 The structure shown is merely a block diagram of a portion of the structure related to the present application and does not constitute a limitation on the computer device to which the present application is applied. Specific computer devices may include more or fewer components than those shown in the figure, or combine certain components, or have different component arrangements.

[0179] In some embodiments of this application, a computer device is provided, including one or more processors; memory; and 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 perform the following steps:

[0180] Obtain the image to be tested of the panel to be tested; the image to be tested includes the light-emitting area corresponding to the light-emitting unit on the panel to be tested;

[0181] Based on the reference luminous region in the image to be detected and the reference luminous region corresponding to the reference luminous region, multiple photometric template images corresponding to different sub-pixel offset values ​​are generated.

[0182] Based on the luminous region corresponding to the luminous unit in the image to be detected and the multiple photometric template images, determine the sub-pixel offset value corresponding to each luminous unit;

[0183] Based on the defect detection results of the image to be detected and the sub-pixel offset values ​​corresponding to each of the light-emitting units, the defective light-emitting units in the panel to be detected are determined.

[0184] In some embodiments of this application, a computer-readable storage medium is provided, storing a computer program that is loaded by a processor, causing the processor to perform the following steps:

[0185] Obtain the image to be tested of the panel to be tested; the image to be tested includes the light-emitting area corresponding to the light-emitting unit on the panel to be tested;

[0186] Based on the reference luminous region in the image to be detected and the reference luminous region corresponding to the reference luminous region, multiple photometric template images corresponding to different sub-pixel offset values ​​are generated.

[0187] Based on the luminous region corresponding to the luminous unit in the image to be detected and the multiple photometric template images, determine the sub-pixel offset value corresponding to each luminous unit;

[0188] Based on the defect detection results of the image to be detected and the sub-pixel offset value corresponding to each of the light-emitting units, the defective light-emitting units in the panel to be detected are determined.

[0189] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium, and when executed, it can include the processes of the embodiments of the methods described above. Any references to memory, storage, databases, or other media used in the embodiments provided in this application can include at least one of non-volatile and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, or optical storage, etc. Volatile memory can include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM can be in various forms, such as static random access memory (SRAM) or dynamic random access memory (DRAM), etc.

[0190] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0191] The panel defect detection method, apparatus, computer device, and storage medium provided in the embodiments of this application have been described in detail above. Specific examples have been used to illustrate the principles and implementation methods of the present invention. The description of the above embodiments is only for the purpose of helping to understand the method and core ideas of the present invention. At the same time, for those skilled in the art, there will be changes in the specific implementation methods and application scope based on the ideas of the present invention. Therefore, the content of this specification should not be construed as a limitation of the present invention.

Claims

1. A method for detecting panel defects, characterized in that, include: Obtain the image to be tested of the panel to be tested; the image to be tested includes the light-emitting area corresponding to the light-emitting unit on the panel to be tested; Based on the reference luminous region in the image to be detected and the reference luminous region corresponding to the reference luminous region, multiple photometric template images corresponding to different sub-pixel offset values ​​are generated. Based on the luminous region corresponding to the luminous unit in the image to be detected and the multiple photometric template images, determine the sub-pixel offset value corresponding to each luminous unit; Based on the defect detection results of the image to be detected and the sub-pixel offset values ​​corresponding to each of the light-emitting units, the defective light-emitting units in the panel to be detected are determined. The step of generating multiple photometric template images corresponding to different sub-pixel offset values ​​based on the reference luminous region in the image to be detected and the corresponding reference luminous region includes: A reference luminous region and a corresponding reference luminous region are determined based on the coordinate information of the target corner points in the image to be detected; the reference luminous region is the region obtained by displacing the reference luminous region. An initial photometric image is generated based on the pixel value difference between corresponding pixels in the reference luminous region and the reference luminous region. The initial photometric image is interpolated based on a preset subpixel offset value to obtain multiple photometric template images corresponding to different subpixel offset values.

2. The panel defect detection method according to claim 1, characterized in that, The step of determining the reference luminous region and the corresponding reference luminous region based on the coordinate information of the target corner point includes: Obtain the width and height of the region corresponding to the light-emitting unit in the image to be detected; Based on the coordinate information of the target corner point and the width and height of the region, the reference luminous area is determined; The reference luminescent region is obtained by translating the reference luminescent region according to the width and height of the region.

3. The panel defect detection method according to claim 1, characterized in that, The reference light-emitting region includes a first reference light-emitting region, a second reference light-emitting region, a third reference light-emitting region, and a fourth reference light-emitting region; The step of generating an initial luminous image based on the pixel value difference between corresponding pixels in the reference luminous region and the reference luminous region includes: The pixel value difference between the reference luminescent region and the corresponding pixels of the first, second, third, and fourth reference luminescent regions is calculated respectively to obtain the first photometric image, the second photometric image, the third photometric image, and the fourth photometric image. An initial photometric image is generated based on the pixel value relationship of corresponding pixels in the first, second, third, and fourth photometric images.

4. The panel defect detection method according to claim 1, characterized in that, The step of determining the sub-pixel offset value corresponding to each light-emitting unit based on the light-emitting region corresponding to the light-emitting unit in the image to be detected and the plurality of photometric template images includes: Calculate the correlation coefficient between the target luminescent region in the image to be detected and each of the photometric template images; Based on the magnitude relationship between the correlation coefficients, the target template image is determined from the photometric template image; The subpixel offset value corresponding to the target template image is determined as the subpixel offset value of the light-emitting unit corresponding to the target light-emitting region.

5. The panel defect detection method according to claim 4, characterized in that, The target luminescent region includes luminescent regions corresponding to multiple target luminescent units; The step of determining the sub-pixel offset value corresponding to the target template image as the sub-pixel offset value of the light-emitting unit corresponding to the target light-emitting region includes: The sub-pixel offset value corresponding to the target template image is determined as the sub-pixel offset value corresponding to the target luminous region; Based on the light-emitting unit corresponding to the target light-emitting region, the sub-pixel offset value corresponding to the target light-emitting region is interpolated to obtain the sub-pixel offset value corresponding to each target light-emitting unit.

6. The panel defect detection method according to any one of claims 1 to 5, characterized in that, The defect detection results include point defect detection results and / or Mura defect detection results; Before determining the defective light-emitting units in the panel to be detected based on the defect detection results of the image to be detected and the sub-pixel offset values ​​corresponding to each of the light-emitting units, the method further includes: Obtain the image to be detected and the corresponding translational image of the image to be detected; Based on the pixel value difference between the corresponding pixels in the translated image and the image to be detected, a residual image of luminance distribution is obtained; Based on the average and variance of the pixel values ​​of each pixel in the residual image of the photometric distribution, abnormal pixels are extracted from the residual image of the photometric distribution. Based on the abnormal pixels, the point defect detection result is determined; and / or, The image to be detected is processed based on a preset point spread function to obtain a photometric distribution image corresponding to the image to be detected; The photometric distribution image is binarized based on the average and variance of the pixel values ​​of each preset region in the photometric distribution image to obtain a binarized image. The Mura defect detection result of the image to be detected is determined based on the connected regions in the binarized image.

7. A panel defect detection device, characterized in that, include: An acquisition module is used to acquire an image of the panel to be inspected; the image to be inspected includes the light-emitting area corresponding to the light-emitting unit on the panel to be inspected; The generation module is used to generate multiple photometric template images corresponding to different sub-pixel offset values ​​based on the reference luminous region in the image to be detected and the reference luminous region corresponding to the reference luminous region. The determining module is used to determine the sub-pixel offset value corresponding to each light-emitting unit based on the light-emitting area corresponding to the light-emitting unit in the image to be detected and the plurality of photometric template images; The detection module is used to determine the defective light-emitting units in the panel to be detected based on the defect detection results of the image to be detected and the sub-pixel offset values ​​corresponding to each of the light-emitting units. The generation module is also used to determine a reference luminous region and a reference luminous region corresponding to the reference luminous region based on the coordinate information of the target corner point in the image to be detected. An initial photometric image is generated based on the pixel value difference between corresponding pixels in the reference luminous region and the reference luminous region. The initial photometric image is interpolated based on a preset subpixel offset value to obtain multiple photometric template images corresponding to different subpixel offset values; the reference luminescent region is the region obtained by displacing the reference luminescent region.

8. 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 steps in the panel defect detection method of any one of claims 1 to 6.

9. A computer-readable storage medium, characterized in that, It stores a computer program, which is loaded by a processor to perform the steps in the panel defect detection method according to any one of claims 1 to 6.

Images