Data processing method and device, storage medium and computer equipment
The uneven brightness defect characteristics of the display screen are extracted through data processing methods, the degree of defect is determined and targeted compensation is carried out, which solves the problem of uneven brightness of the display screen, improves product yield and efficiency, and reduces costs.
Patent Information
- Application Number
- CN202310549186.5
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2023-05-12
- Publication Date
- 2024-11-12
AI Technical Summary
During the display production process, the uneven brightness phenomenon (Mura) caused by material and process reasons causes the product to fail to meet the specification requirements of end customers, resulting in downgrading or scrapping. The existing technology is difficult to detect, frequent rework, and low efficiency.
By using the data processing method, by obtaining the optical image data of the display screen, visual features are extracted, the degree of defect is judged, and targeted compensation processing is carried out based on the degree of defect information, reducing the number of compensation, improving the yield of one time, and reducing the rework cost.
It improves the yield of the display screen, reduces the number of rework times, reduces labor costs, improves the accuracy of defect degree information and the accuracy of compensation processing, and avoids the reduction in yield caused by multiple compensations.
Smart Images

Figure CN118941837A_ABST
Abstract
Description
Technical Field
[0001] The present application relates to the technical field of display screen defects, and in particular to a data processing method, device, storage medium and computer equipment. Background Art
[0002] During the display production process, due to reasons such as materials and processes, some products may have uneven screen brightness, which is called Mura. These uneven brightness spots will cause visual discomfort and reduce the yield of the display. Products with such spots cannot meet the specifications of end customers and can only be scrapped or downgraded. Summary of the invention
[0003] The embodiments of the present application provide a data processing method, an apparatus, a storage medium, and a computer device, which can use optimal defect compensation parameters to compensate for display screen defects and improve the yield of the display screen.
[0004] The present application provides a data processing method, including:
[0005] Acquiring optical image data of the display screen;
[0006] Performing visual feature extraction processing on the optical image data according to the defect type of the display screen defect to obtain a defect feature set corresponding to at least one defect type, wherein each defect type corresponds to a defect feature set;
[0007] Determine and process the defect features in each corresponding defect feature set to obtain defect degree information corresponding to at least one defect type;
[0008] Compensation processing is performed on display screen defects of corresponding defect types according to the defect degree information.
[0009] The present application provides a data processing device, including:
[0010] An image acquisition module, used to acquire optical image data of the display screen;
[0011] A feature extraction module, configured to perform visual feature extraction processing on the optical image data according to the defect type of the display screen defect, so as to obtain a defect feature set corresponding to at least one defect type, wherein each defect type corresponds to a defect feature set;
[0012] A determination module, used for performing determination processing on the defect features in each corresponding defect feature set to obtain defect degree information corresponding to at least one defect type;
[0013] The compensation module is used to compensate display screen defects of corresponding defect types according to the defect degree information.
[0014] An embodiment of the present application further provides a computer-readable storage medium, in which a plurality of instructions are stored, and the instructions are suitable for being loaded by a processor to execute the steps in any of the above methods.
[0015] An embodiment of the present application further provides a computer device, including a processor and a memory, wherein the processor is electrically connected to the memory, the memory is used to store instructions and data, and the processor is used to perform the steps in any of the methods described above.
[0016] The data processing method, device, storage medium and computer equipment provided by the present application can determine the defect degree information corresponding to the defect type of the display screen, which is equivalent to classifying the defect degree of the defect type of the display screen defect, rather than classifying the defect type. Visual feature extraction processing is performed on the optical image data of the display screen according to the defect type of the display screen defect to obtain a defect feature set corresponding to at least one defect type, and the defect features in each corresponding defect feature set are determined to obtain defect degree information corresponding to at least one defect type, thereby improving the accuracy of determining the defect degree information. After obtaining the defect degree information, the display screen defect corresponding to the defect type can be compensated according to the defect degree information. Using more accurate defect degree information to compensate for the display screen defects can not only reduce the number of compensation processing times in a targeted manner, but also improve the one-time yield of the display screen, avoid multiple compensations causing a reduction in the yield of the display screen, and simultaneously avoid multiple rework, that is, reduce rework and reduce labor costs. BRIEF DESCRIPTION OF THE DRAWINGS
[0017] The technical solution and other beneficial effects of the present application will be made apparent by describing in detail the specific implementation methods of the present application in conjunction with the accompanying drawings.
[0018] Figure 1 It is a schematic diagram of the Demura process in the prior art.
[0019] Figure 2 This is an example flowchart of the data processing method provided in an embodiment of the present application.
[0020] Figure 3 A schematic diagram of corresponding defect scores and optimal defect compensation parameters provided in an embodiment of the present application.
[0021] Figure 4 A simplified flowchart of a data processing method provided in an embodiment of the present application.
[0022] Figure 5 A flowchart of a method for training a defect degree determination model provided in an embodiment of the present application.
[0023] Figure 6 A schematic diagram of a sub-process of a method for training a defect degree determination model provided in an embodiment of the present application.
[0024] Figure 7 A simplified flowchart of a method for training a defect degree determination model provided in an embodiment of the present application.
[0025] Figure 8 A schematic diagram showing the brightness unevenness score provided in an embodiment of the present application.
[0026] Fig. 9 A schematic diagram showing the color unevenness score provided in an embodiment of the present application.
[0027] Fig.10 Schematic diagram of the accuracy of data sets on different production lines.
[0028] Fig.11 A schematic diagram of the structure of a data processing device provided in an embodiment of the present application.
[0029] Fig.12 A structural diagram of the training of the defect degree determination model provided in an embodiment of the present application.
[0030] Fig.13 A schematic diagram of the structure of a computer device provided in an embodiment of the present application.
[0031] Fig.14 Another structural schematic diagram of a computer device provided in an embodiment of the present application. DETAILED DESCRIPTION
[0032] The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative work are within the scope of protection of this application.
[0033] The embodiments of the present application provide a data processing method, device, storage medium and computer equipment. Any method and / or any device provided in the embodiments of the present application can be integrated in a computer device, which includes any device that can realize the determination of the defect degree information of the Mura defect of the display screen, and further, any compensation device that can compensate for the Mura defect of the display screen according to the defect degree information. The computer device in the present application can be an independent device or a device cluster, such as determining the defect degree information of the Mura defect of the display screen and further compensating for the Mura defect of the display screen according to the defect degree information. The steps of determining the defect degree information of the Mura defect of the display screen and further compensating for the Mura defect of the display screen according to the defect degree information are performed in one computer device, and the training to obtain the defect degree determination model is performed in another computer device, and of course, it can also be integrated in one computer device for execution.
[0034] Before introducing the solution in the embodiment of the present application in detail, the existing Demura process is briefly introduced. In the existing Demura process, all display products use the same Demura process such as the main line process (main line code). Accordingly, all display products use the same compensation parameters. When the display quality is poor, the rework process (Rework code) is used. The rework process will increase additional costs. When it is detected that the display quality meets the requirements, it will be shipped. Figure 1 As shown. The existing Demura process requires manual determination of the fault type and repeated rework, which wastes manpower and is inefficient. At present, Mura is usually detected as point and line defects, but the point and line features of the common Mura defects in the actual production process are not obvious, which is not conducive to the current Mura defect detection and it is difficult to solve the problem of the technical production line.
[0035] The data processing method, apparatus, computer-readable storage medium and computer device in the present application are described in detail below, wherein the order of description of the embodiments below is not intended to limit the priority order of the embodiments.
[0036] Figure 2 It is a flow chart of the data processing method provided in the embodiment of the present application, and the method comprises the following steps.
[0037] 101, acquiring optical image data of a display screen.
[0038] The display screen is photographed by a computer device such as an industrial camera of a compensation device to obtain optical image data of the display screen, wherein the industrial camera has higher accuracy than ordinary cameras.
[0039] 102 . Perform visual feature extraction processing on the optical image data according to the defect type of the display screen defect to obtain a defect feature set corresponding to at least one defect type, wherein each defect type corresponds to a defect feature set.
[0040] The defect types of display screen defects include uneven color defects, uneven brightness defects or other defects. For example, different defect features are first designed for display screen defects of different defect types, and then visual feature extraction processing is performed on the optical image data using visual feature extraction algorithms corresponding to different defect features according to the defect types to obtain defect feature sets corresponding to different defect types, wherein different defect features of each defect type constitute a defect feature set, that is, one defect type corresponds to one defect feature set, and one defect feature set includes at least one defect feature. According to this method, a defect feature set corresponding to at least one defect type can be obtained.
[0041] For example, the color unevenness defect includes but is not limited to at least one of a color feature and a saturation feature, wherein the color feature refers to a local color feature and / or a global color feature of the color unevenness defect, and the saturation feature refers to a local saturation feature and / or a global saturation feature of the color unevenness defect.
[0042] Among them, the color characteristics of color unevenness defects can be obtained through the following visual feature extraction algorithm: converting the optical image data of the display screen from the RGB space into other color spaces, such as HSV (Hue, Saturation and Value) color space, YUV color space, HIS (Hue, Intensity, Saturation) color space, etc.; dividing the converted optical image data into blocks, for example, into a plurality of blocks arranged horizontally and vertically, such as nine blocks arranged three horizontally and three vertically; determining the color difference between each block and the central block to obtain a global color difference feature, determining the color difference between each block and the surrounding blocks to obtain a local color difference feature, using the global color difference feature as the global color difference feature, and using the local color difference feature as the local color feature.
[0043] Among them, the saturation feature of the color unevenness defect can be obtained by the following visual feature extraction algorithm: converting the optical image data of the display screen from the RGB space to other color spaces, such as HSV (Hue, Saturation and Value) color space, YUV color space, HIS (Hue, Intensity, Saturation) color space, etc.; dividing the converted optical image data into blocks, for example, into a plurality of blocks arranged horizontally and vertically, such as nine blocks arranged three horizontally and three vertically; determining the saturation difference between adjacent blocks to obtain a local saturation feature, and determining the difference mean of the saturation difference of multiple blocks to obtain a global saturation feature.
[0044] For example, the defect features corresponding to the uneven brightness defect include but are not limited to at least one of the features of light and dark, horizontal and vertical stripes, overall brightness, local brightness and messy stripes. Among them, the light and dark feature refers to the difference feature between the bright part and the dark part in the optical image data, the horizontal and vertical stripes feature refers to the vertical stripe feature and the horizontal stripe feature obtained after horizontal and vertical projection of the optical image data, the overall local brightness feature refers to the feature related to the brightness information in all blocks after the optical image data is divided into blocks, the local brightness adjustment refers to the feature related to the gradient feature information in all blocks after the optical image data is divided into blocks, and the messy stripe feature refers to the feature related to the stripe area corresponding to the gradient obvious area in the optical image data. Among them, when calculating the defect features corresponding to the uneven brightness defect, except for the horizontal and vertical stripes feature which only takes one channel data, when processing the others, it is necessary to perform the corresponding processing below on the three channels of RGB or convert the RGB optical image data to grayscale, and then perform the corresponding processing below on the grayscale image after the grayscale processing, which will not be repeated below.
[0045] Among them, the light and dark features of the uneven brightness defect can be obtained by the following visual feature extraction algorithm: Gaussian blur processing is performed on the optical image data, and edge detection processing is performed on the image data after Gaussian blur processing to obtain a gradient edge map; the gradient edge map is segmented using a segmentation algorithm according to different brightness levels in the gradient edge map to obtain a segmentation result; the noise blocks in the segmentation result are removed using an image morphological method to obtain light and dark area blocks, which include light area blocks and dark area blocks; the area difference of the light and dark area blocks and the brightness difference between the light and dark area blocks are determined; the difference coefficient of the light and dark area blocks is determined using the area difference and the difference difference, for example, the difference coefficient of the light and dark area blocks = the area difference of the light and dark area blocks * the brightness difference between the light and dark area blocks; the difference coefficient of the light and dark area blocks is determined as the light and dark feature. The light and dark feature is a numerical value.
[0046] Among them, the horizontal and vertical stripes features of uneven brightness defects can be obtained by the following visual feature extraction algorithm: obtain any channel data in the optical image data, such as the G channel data, and perform horizontal and vertical projections on the G channel data to obtain horizontal and vertical components. For example, for 1080*780 G channel data, add the data of each row to obtain the horizontal component, then the horizontal component corresponds to 1080 data, add the data of each column to obtain the vertical component, then the vertical component corresponds to 780 data, determine the gradient features of the horizontal component and the vertical component, use the gradient features of the horizontal component as the horizontal stripe feature, and use the gradient features of the vertical component as the vertical stripe feature; determine the number of dark areas represented in the horizontal stripe features and the vertical stripe features according to the first gradient feature threshold, and determine the final horizontal and vertical stripe features according to the number of dark areas, such as using the number of dark areas as the final horizontal and vertical stripe features, and the horizontal and vertical stripe features correspond to a quantitative feature.
[0047] Among them, the overall brightness characteristics of uneven brightness defects can be obtained through the following visual feature extraction algorithm: divide the optical image data into blocks, for example, divide it into rows and columns, and divide it into multiple blocks arranged horizontally and vertically; calculate the average brightness of each block after block division; determine the maximum brightness difference in the corresponding average brightness between all blocks, and the brightness variance of the corresponding average brightness between all blocks; determine the overall brightness characteristics based on the maximum brightness difference and the brightness variance.
[0048] Among them, the local brightness features of uneven brightness defects can be obtained through the following visual feature extraction algorithm: divide the optical image data into blocks, for example, divide them into blocks according to rows and columns, and divide them into multiple blocks arranged horizontally and vertically; calculate the average brightness of each block after block division; determine the gradient features between the average brightness of each block and the average brightness of the surrounding blocks as local difference features; use the second gradient threshold to process the determined gradient features to distinguish between transition-obvious areas and transition-smooth areas, and give different weights to different areas; use the different weights of local difference features, transition-obvious areas, and transition-smooth areas as local brightness features.
[0049] The messy stripe feature of the uneven brightness defect can be obtained by the following visual feature extraction algorithm: performing denoising on the optical image data to remove noise, performing gradient calculation on the optical image data after the noise is removed to obtain a gradient map, screening the gradient map to obtain an area with obvious gradient, that is, an area with a large gradient value, and taking the area with obvious gradient as the messy stripe area; determining the brightness difference between the messy stripe area and the surrounding area, and the stripe area of the messy stripe area, and determining the messy stripe feature according to the brightness difference and the stripe area, for example, the messy stripe feature = the brightness difference between the stripe and the surrounding area * the stripe area; and / or, quantifying the point and line features of the messy stripe area, that is, determining the SEMU feature of the messy stripe area, and taking the SEMU feature as the messy stripe feature, wherein the SEMU feature can be obtained using the formula To determine, where |C x | represents the brightness difference between the messy stripe area and the surrounding area, s x Represents the stripe area.
[0050] It should be noted that the above visual feature extraction algorithm is only an example, and other visual feature extraction algorithms can also be used to achieve it. It should also be noted that the defect features of uneven brightness and uneven color defects designed above are obtained by combining countless experiences of experts in judging Mura defects, and these defect features can also be transferred to other defect types.
[0051] For example, for color unevenness defects, visual feature extraction processing is performed on the optical image data according to a visual feature extraction algorithm of color features and / or saturation features of the color unevenness defects to obtain a defect feature set of the color unevenness defects, which includes color features and / or saturation features.
[0052] For example, for uneven brightness defects, the optical image data is processed for world feature extraction according to the corresponding visual feature extraction algorithm to obtain a defect feature set of color uneven defects, which includes but is not limited to at least one defect feature among light and dark features, horizontal and vertical stripe features, overall brightness features, local brightness features and messy stripe features.
[0053] 103 , performing determination processing on the defect features in each corresponding defect feature set to obtain defect degree information corresponding to at least one defect type.
[0054] The defect features in the corresponding defect feature set may be normalized to obtain a normalized feature set; and each normalized feature in the normalized feature set may be subjected to degree determination to obtain defect degree information corresponding to the defect type.
[0055] Among them, the defect degree information represents the severity information of the display screen defect corresponding to the defect type, wherein the defect degree information can be represented by a defect score, for example, the higher the defect score, the more serious the defect degree; it can also be represented by information representing the level such as medium, high and low, where high corresponds to the most serious defect degree, and low corresponds to the least serious defect degree, etc.
[0056] In some cases, a model can be used to determine the defect degree information corresponding to a defect type. For example, a defect degree determination model that matches the defect type is used to perform determination processing on the defect features in each corresponding defect feature set to obtain defect degree information corresponding to at least one defect type, wherein the defect degree determination model that matches the defect type is obtained based on learning the defect degree sorting order information corresponding to the defect types of multiple display screens.
[0057] Among them, one defect type may correspond to one defect degree determination model, or one defect degree determination model may be used to determine multiple defect types to obtain defect degree information of multiple defect types. When one defect type corresponds to one defect degree determination model, it is necessary to call the defect degree determination model that matches the defect type to determine the defect type currently. When one defect degree determination model can determine multiple defect types, the defect degree determination model that matches the defect type refers to the defect degree determination model. In the embodiment of the present application, one defect type corresponds to one defect degree determination model as an example for explanation.
[0058] Assuming that it is necessary to determine the severity information of the color unevenness defect of the display screen defect, the defect degree determination model corresponding to the color unevenness defect is called, and the defect features in the defect feature set corresponding to the color unevenness defect are input into the corresponding defect degree determination model for determination processing to obtain the defect degree information corresponding to the color unevenness defect. If it is necessary to determine the severity information of the brightness unevenness defect of the display screen defect, the defect degree determination model corresponding to the brightness unevenness defect is called, and the defect features in the defect feature set corresponding to the brightness unevenness defect are input into the corresponding defect degree determination model for determination processing to obtain the defect degree information corresponding to the brightness unevenness defect.
[0059] Among them, the defect degree determination model includes a normalization module and a degree determination module. The normalization module is used to normalize the defect features in the corresponding defect feature set to obtain a normalized feature set. The degree determination module is used to perform degree determination processing on each normalized feature in the normalized feature set to obtain defect degree information corresponding to the defect type.
[0060] It is assumed that each defect feature set is represented by Feats, and each defect feature in Feats is represented by feats.i , i=1…k is used to represent it.
[0061] In one embodiment, a normalization module may be used to normalize each defect feature in the defect feature set to obtain a normalized feature in the normalized feature set. Specifically, the following formula (1) can be used to obtain the normalized features.
[0062]
[0063] in, Refers to the minimum value of a single defect feature in all displays, such as the minimum value of the defect feature in all displays in the optical image data set during the model training phase. Refers to the maximum value of a single defect feature in all displays, such as the maximum value of the defect feature in all displays in the optical image dataset during the model training phase. i Represents the i-th defect feature in the defect feature set. Since the maximum and minimum values in the optical image dataset are used during training, the same maximum and minimum values are also used during the model use phase to improve accuracy.
[0064] In other embodiments, the softmax network layer or other methods may be directly used to normalize each defect feature in the defect feature set.
[0065] After the normalized features are obtained, the degree determination module may be used to perform degree determination processing on each normalized feature to obtain defect degree information corresponding to the defect type.
[0066] In one embodiment, a network module in a degree determination module is used to perform degree determination processing on each normalized feature to obtain a defect degree probability corresponding to the defect type, and defect degree information is determined based on the defect degree probability, such as directly using the defect degree probability as the defect degree information, or determining the defect degree information based on the defect degree probability according to a first preset rule, such as determining 0-15% as a low defect level.
[0067] In one embodiment, the degree determination module can also obtain the weight corresponding to each normalized feature, and the weight corresponding to each normalized feature can be obtained according to the training defect degree determination model described below. The process of training the defect degree determination model can also be understood as the process of obtaining the weight corresponding to each normalized feature, and the defect degree determination processing is performed on each normalized feature according to the weight to obtain the defect score corresponding to the defect type, and the defect degree information is determined according to the defect score, such as directly using the defect score as the defect degree information. This case is used as an example in the embodiment of the present application, or the defect degree information is determined according to the defect score according to the second preset rule, such as determining a defect level of 0-10 as a low defect level, and a defect score of 20-35 as a medium-low defect level, etc.
[0068] Specifically, formula (2) can be used to obtain the corresponding defect degree information.
[0069]
[0070] Among them, w i It represents the weight corresponding to the i-th normalized feature, and score represents the defect score (defect degree information).
[0071] It should be noted that no matter how the normalization module and the degree determination module of the defect degree determination model are implemented, the defect degree determination model in the parameter determination stage (training stage) is obtained based on learning the defect degree sorting order information corresponding to the defect types of multiple display screens (corresponding to the defect degree sorting label information in the training stage), that is, in the parameter determination stage, the defect degree label information of each display screen in the multiple display screens, such as the defect score, is not learned, but the defect degree of each display screen in the multiple display screens is sorted to obtain the defect degree sorting order information of the multiple display screens, and finally the corresponding parameters of the defect degree determination model are obtained according to the defect degree sorting order information of the multiple display screens.
[0072] In the embodiment of the present application, a defect degree determination model is creatively used to fit the defect degree sorting order of multiple display screens. This method of obtaining the corresponding parameters of the defect degree determination model based on the defect degree sorting order information of multiple display screens is, on the one hand, due to the fact that defect determination has many ambiguities, it is very difficult to directly give the defect degree information of each display screen, such as the defect score, but it is much simpler to sort the defect degrees of multiple display screens. On the other hand, learning the defect degree sorting order information of multiple display screens avoids overfitting caused by directly using defect scores to train the model, which affects the robustness of the defect degree determination model. Learning the defect degree sorting order information of multiple display screens makes the defect degree determination model obtained more robust and more generalized, and improves the accuracy of the defect degree information corresponding to the defect type of the display screen. In addition, in the present application, a defect degree determination model matching the defect type is used to classify the defect degree, so that the defect degree information obtained is objective and accurate.
[0073] In one embodiment, part of the visual feature extraction processing may also be implemented by a defect degree determination model, that is, part of the visual feature extraction processing is integrated into the defect degree determination model, and the specific implementation principle remains unchanged and will not be repeated here.
[0074] 104 , performing compensation processing on display screen defects of corresponding defect types according to the defect degree information.
[0075] Since the defect degree information represents the severity of the defect, the display screen defects of the corresponding defect type can be compensated more specifically according to the defect degree information.
[0076] Since the present application can improve the accuracy of defect degree information, compensation processing is performed on display screen defects of corresponding defect types directly according to the defect degree information to improve the accuracy of compensation processing. Using the defect degree information to compensate for display screen defects can not only reduce the number of compensation processing times in a targeted manner, but also improve the one-time yield of the display screen, avoid multiple compensations causing a reduction in the yield of the display screen, and at the same time avoid multiple rework, that is, reduce rework and reduce labor costs.
[0077] In one embodiment, step 104 includes: obtaining optimal defect compensation parameters corresponding to the defect degree information, and performing compensation processing on display screen defects corresponding to corresponding defect types according to the optimal defect compensation parameters.
[0078] During the debugging process, after using the defect degree determination model to obtain the defect degree information corresponding to the defect type, such as the defect score, multiple score intervals or multiple defect levels are divided according to the defect degree information corresponding to the defect type, such as the defect score, and the optimal defect compensation parameters, that is, the optimal Demura algorithm parameters, are debugged for each score interval / defect level.
[0079] It should be noted that the premise for doing this is that, through countless display screen cases in the production process, we found that the severity of the display screen's optical image data is correlated with the difficulty of Demura. Therefore, the optimal defect compensation parameters can be debugged specifically according to the defect score.
[0080] In actual use, after obtaining defect degree information such as a defect score corresponding to the defect type of the display screen, the optimal defect compensation parameter corresponding to the defect degree information such as the defect score is obtained.
[0081] like Figure 3 As shown, in the debugging stage, for multiple groups of optical image data of several display screens, defect degree information corresponding to the defect type, such as defect score, is obtained, and then multiple score intervals are divided according to the defect score, and debugging is performed for each score interval to obtain the optimal defect compensation parameters of the corresponding score interval. In specific use, that is, in the production stage, the optical image data obtained by taking a photo of the display screen is obtained, the defect degree information corresponding to the defect type, such as defect score, is obtained, and the optimal defect compensation parameters of Demura corresponding to the defect score are obtained.
[0082] When the display screen defect of the display screen has a defect type and corresponding defect degree information is obtained, the optimal defect compensation parameter corresponding to the defect degree information can be directly obtained.
[0083] Wherein, when the display screen defect of the display screen has multiple defect types and correspondingly obtains multiple defect degree information, the above step of obtaining the optimal defect compensation parameter corresponding to the defect degree information includes: determining the target defect degree information from the defect degree information corresponding to the multiple defect types, and obtaining the optimal defect compensation parameter corresponding to the target defect degree information. Wherein, if the defect degree information corresponding to the multiple defect types is different, then the defect degree information with the highest defect degree information is obtained from the defect degree information corresponding to the multiple defect types as the target defect degree information; if the defect degree information corresponding to the multiple defect types is the same, then the defect degree information of the preset defect type is obtained from the defect degree information corresponding to the multiple defect types as the target defect degree information. Wherein, the preset defect type can be determined in advance, for example, uneven brightness defect is used as the preset defect type. Multiple preset defect types can be set, or different defect types can be set each time.
[0084] For example, there are two defect types, namely, uneven brightness defect and uneven color defect, among which, the defect score of uneven brightness defect is 0.2, and the defect score of uneven color defect is 0.5, then the defect degree information of uneven color defect is the highest, and the defect degree information of uneven color defect is used as the target defect degree information; if the defect scores of uneven brightness defect and uneven color defect are both 0.5, the defect degree information of the preset defect type such as uneven brightness defect is used as the target defect degree information.
[0085] After determining the target defect degree information, the optimal defect compensation parameters corresponding to the target defect degree information can be directly obtained. For example, after determining that the defect degree information of the uneven brightness defect is the target defect degree information, the optimal defect compensation parameters corresponding to the defect degree information of the uneven brightness defect can be directly obtained.
[0086] In one embodiment, when there are multiple defect types and correspondingly multiple defect degree information, the corresponding optimal defect compensation parameters can be obtained for each defect degree information to achieve compensation processing for the display screen defects corresponding to each defect degree information; or multiple defect degree information are determined and processed, and when the defect degree information is lower than a certain threshold, such as the defect score is lower than a certain threshold, or the defect level is low, the optimal defect compensation parameters corresponding to the defect degree information are no longer obtained, so that the display screen defects corresponding to the defect degree information are not compensated, and the corresponding optimal defect compensation parameters are obtained for the defect degree information higher than a certain threshold to compensate for the corresponding display screen defects.
[0087] After obtaining the optimal defect compensation parameters, compensation processing is performed on the display screen defects corresponding to the corresponding defect type according to the optimal defect compensation parameters.
[0088] For example, if it is an optimal defect compensation parameter corresponding to the uneven brightness defect, the uneven brightness defect is directly compensated according to the optimal defect compensation.
[0089] The display screen defects corresponding to the defect type are compensated directly according to the optimal defect compensation parameters. This can not only reduce the number of compensation processing times in a targeted manner, but also improve the first-time yield of the display screen, avoid the reduction of the display screen yield due to multiple compensations, and avoid multiple rework at the same time, that is, reduce rework.
[0090] like Figure 4As shown, it is a simple flow chart of the data processing method provided by the embodiment of the present application. Before the display screen defects are compensated, an industrial camera is used to take an optical photo relative to the display screen to obtain optical image data, and then the optical image data is subjected to visual feature extraction processing, and then the defect feature set corresponding to at least one defect type obtained by the visual feature extraction processing is input into the defect degree determination model to obtain defect degree information such as defect score, and further determine the defect severity according to the defect score. For example, if the color unevenness is high (the defect severity of the color unevenness defect is high), the optimal defect compensation parameter corresponding to the high color unevenness is called to compensate the color unevenness defect, and if the brightness unevenness is high (the defect severity of the brightness unevenness defect is high), the optimal defect compensation parameter corresponding to the high brightness unevenness is called to compensate the brightness unevenness defect; when the color unevenness is the lowest or the brightness unevenness is the lowest, the optimal defect compensation parameter corresponding to the lowest time can be called for compensation, or compensation can be performed directly according to the main line process. If there is still a problem, the parameters in the main line process (correction code) are corrected. After the compensation process is completed, the display screen can be shipped.
[0091] In this embodiment, when performing Demura, the severity of defects of the display screen before Demura is taken into consideration, so as to optimize the Demura process, improve the first-time yield, and reduce rework.
[0092] The above is described from the perspective of data processing method. The following will describe the training method of the defect degree determination model, which is used to train the defect degree determination model mentioned above.
[0093] like Figure 5 As shown, a training method for a defect degree determination model is provided, and the training method includes the following steps.
[0094] 201. Obtain an optical image data set, wherein the optical image data set includes a plurality of training optical image data of a plurality of display screens corresponding to each defect type, the plurality of training optical image data are divided into a plurality of groups, and the plurality of training optical image data in each group correspond to defect degree ranking label information.
[0095] The optical image data set includes a plurality of training optical image data of multiple display screens corresponding to each defect type, that is, each defect type includes a plurality of training optical image data of multiple display screens. One display screen may correspond to one piece of training optical image data. If multiple different defects appear in one display screen, one piece of training optical image data may correspond to each of the different defect types.
[0096] For each defect type, the several training optical image data corresponding to the defect type are divided into multiple groups, and the multiple training optical image data in each group correspond to defect degree ranking label information, that is, the defect degree ranking order information mentioned above. Among them, the defect degree ranking label information in the optical image data set can be the defect degree ranking order determined by the Demura expert, or the defect degree ranking order determined by other methods.
[0097] In one embodiment, for each defect type, the following steps are followed to train a defect degree determination model that matches the defect type. This embodiment of the present application takes this case as an example for illustration. In another embodiment, an optical image data set can be used as input for training to obtain a defect degree determination model that can determine multiple defect types.
[0098] 202 , performing visual feature extraction processing on the multiple training optical image data in each group to obtain a training defect feature set corresponding to the defect type.
[0099] Visual feature extraction is performed on multiple training optical image data in each group according to the defect type, wherein different defect types correspond to different visual feature extraction algorithms, and the proposed training defect features are different. Each training optical image data corresponds to a training defect feature set, wherein the specific visual feature extraction algorithm is described above and will not be repeated here.
[0100] 203 , using the initial defect degree determination model, performing determination processing on the training defect feature set to obtain defect degree prediction information of the defect type.
[0101] The initial defect degree determination model refers to a model whose parameters need to be changed, and the initial defect degree determination model includes a normalization module and a degree determination module.
[0102] The normalization module is used to normalize the training defect features in the training defect feature set to obtain the training normalized feature set; the program determination module is used to perform degree determination processing on each training normalized feature in the training normalized feature set to obtain defect degree prediction information corresponding to the defect type.
[0103] Among them, the specific method of normalizing the training defect feature set obtained from each training optical image data can be referred to as shown in formula (1) above, and then the weight (initial weight) of each training normalized feature is set, and the judgment processing is performed according to the weight (initial weight) and the corresponding training normalized feature to obtain the defect degree prediction information corresponding to the defect type, wherein the defect degree prediction information corresponding to the defect type can be obtained as shown in formula (2) above, and the details are not repeated here. The corresponding prediction degree prediction information includes the defect prediction score.
[0104] 204 , determining defect degree ranking prediction information corresponding to the plurality of training optical image data in each group according to the defect degree prediction information.
[0105] Corresponding defect level prediction information can be obtained for multiple training optical image data in each group, so that multiple defect level prediction information is obtained, and the multiple defect level prediction information, such as multiple defect prediction scores, are sorted to obtain defect level ranking prediction information corresponding to the multiple training optical image data in each group. The sorting rules corresponding to the sorting process are the same as the sorting rules corresponding to the defect level ranking label information, such as sorting from large to small.
[0106] 205 , updating the parameters of the initial defect level determination model according to the defect level ranking prediction information and the defect level ranking label information in each group, so as to obtain a defect level determination model matching the defect type.
[0107] Among them, Figure 6 As shown, step 205 includes the following steps 2051 to 2053.
[0108] 2051, determine the distance between the defect level ranking prediction information and the defect level ranking label information in each group.
[0109] Assume that the defect degree sorting label information in each group is Order i To represent, the defect degree ranking prediction information is used to Indicates, where i = 1…n, assuming that there are n groups, each group includes m training optical image data. The Order can be determined according to the following formula (3): i and The distance between.
[0110]
[0111] Among them, d i It represents the distance between the defect level ranking label information and the defect level ranking prediction information in each group, or it can also be understood as the order difference between the defect level ranking label information and the defect level ranking prediction information.
[0112] 2052, determine the similarity between the defect level ranking prediction information and the defect level ranking label information in each group according to the distance.
[0113] In one embodiment, the similarity may be represented by the Spearman correlation coefficient. In other embodiments, other similarities may be used to represent the similarity.
[0114] The Spearman correlation coefficient can be used to express the similarity according to the following formula (4).
[0115]
[0116] Among them, Corre i represents the Spearman correlation coefficient, m refers to the number of training optical image data in each group. If one display screen corresponds to one training optical image data, then m is also the number of display screens in each group.
[0117] 2053, using the parameters corresponding to the similarity that meets the preset conditions to update the parameters of the initial defect degree determination model, so as to obtain a defect degree determination model that matches the defect type.
[0118] When the similarity includes the Spearman correlation coefficient, the preset condition is determined to be met when the mean of the Spearman correlation coefficient in each group is maximum, and the parameter when the mean of the Spearman correlation coefficient in each group is maximum is used as the parameter corresponding to the similarity that meets the preset condition.
[0119] There are n groups in total, and the mean of the Spearman correlation coefficient in each group can be expressed using formula (5).
[0120]
[0121] The mean of the Spearman correlation coefficient in each group can be expressed using the maximum formula (6).
[0122]
[0123] Among them, the search weight makes formula (6) as large as possible. The search algorithm for the search weight can use random search, brute force search and heuristic search algorithm (such as ant colony algorithm, genetic algorithm, simulated annealing algorithm, etc.). Among them, after a large number of experiments, it is found that the iterative heuristic combined with linear search has the best effect. After training, the weight obtained by formula (6) is the parameter of the defect degree judgment model.
[0124] This training process converts expert experience into computer vision defect features feat and the weight w of each defect feature.
[0125] In other embodiments, other methods may be used to obtain the parameters of the network of the defect degree determination module. Among them, the loss value is determined according to the defect degree ranking prediction information and defect degree ranking label information in each group, and the parameters of the initial defect degree determination model are updated according to the loss value. When certain conditions are met, such as the number of training rounds reaches a preset number of rounds, the model converges, or the loss value converges, the training is stopped, and the obtained model is the defect degree determination model.
[0126] The above training method uses defect degree sorting order information to make the defect degree information obtained by the defect degree determination model, such as defect score, have high accuracy and a wide range of applications. The determination of defect degree information can be transferred to different production lines with high generalization.
[0127] like Figure 7 It is a simple schematic diagram of the entire training process provided by the embodiment of the present application. After obtaining n groups of multiple training optical image data corresponding to a certain defect type, each group of multiple training optical image data is visualized, and the multiple training optical image data in each group are subjectively sorted by Demura experts to obtain defect degree sorting label information. Then, the multiple training optical image data in each group are preprocessed, such as grayscale processing, and visual feature extraction processing is performed using a visual feature extraction algorithm. The results of the visual feature extraction processing are then judged using a defect degree judgment model to obtain defect degree prediction information corresponding to the defect type, such as a defect prediction score. The multiple training optical image data in each group are sorted according to the defect prediction score to obtain defect degree sorting prediction information for each group, and the defect sorting prediction information (algorithm sorting) is compared with the defect degree sorting label information (expert sorting) to make the algorithm sorting consistent with the expert sorting.
[0128] Among them, Figure 8 and Fig. 9 As shown, the defect scores of uneven brightness and uneven color are displayed respectively. The defect scores are the defect scores obtained by the defect degree determination model. It can be seen that the algorithm ranking is consistent with the expert ranking.
[0129] After the defect degree determination model is obtained, the accuracy is very high on the test set and the validation set, meeting the acceptance criteria. Among them, part of the optical image data set in production line 1 is used as a training set, such as 80% as a training set, and the other part is used as a test data set, such as 20% as a test data set. The optical image data set in production line 2 is other production lines and is only used as a test data set. Fig.10 It can be seen that the experience of the defect degree determination model in the embodiment of the present application in determining the defect degree information of the defect type can be transferred, the defect degree determination model is generalizable, and the features are also generalizable.
[0130] According to the method described in the above embodiment, this embodiment will be further described from the perspective of a device. The device can be implemented as an independent entity or integrated in a computer device.
[0131] See also Fig.11 , Fig.11 The data processing device provided in the embodiment of the present application is specifically described, and is applied in a computer device. The data processing device 300 may include: an image acquisition module 301 , a feature extraction module 302 , a determination module 303 and a compensation module 304 .
[0132] The image acquisition module 301 is used to acquire optical image data of the display screen.
[0133] The feature extraction module 302 is used to perform visual feature extraction processing on the optical image data according to the defect type of the display screen defect to obtain a defect feature set corresponding to at least one defect type, wherein each defect type corresponds to a defect feature set.
[0134] The determination module 303 is used to perform determination processing on the defect features in each corresponding defect feature set to obtain defect degree information corresponding to at least one defect type. For example, it is used to perform determination processing on the defect features in each corresponding defect feature set using a defect degree determination model matching the defect type to obtain defect degree information corresponding to at least one defect type, wherein the defect degree determination model matching the defect type is obtained based on learning defect degree sorting order information corresponding to the defect type of multiple display screens.
[0135] In one embodiment, the defect degree determination model includes a normalization module and a degree determination module. The model determination module 303 is specifically used to use the normalization module to normalize the defect features in the corresponding defect feature set to obtain a normalized feature set; and use the degree determination module to perform degree determination processing on each normalized feature in the normalized feature set to obtain the defect degree information corresponding to the defect type.
[0136] The compensation module 304 is used to perform compensation processing on display screen defects of corresponding defect types according to the defect degree information.
[0137] Among them, the compensation module includes a parameter acquisition module and a defect compensation module.
[0138] Among them, the parameter acquisition module is used to obtain the optimal defect compensation parameters corresponding to the defect degree information.
[0139] When the display screen defect has multiple defect types and corresponds to multiple defect degree information, the parameter acquisition module is specifically used to determine the target defect degree information from the defect degree information corresponding to the multiple defect types; and obtain the optimal defect compensation parameter corresponding to the target defect degree information. Wherein, determining the target defect degree information from the defect degree information corresponding to the multiple defect types includes: if the defect degree information corresponding to the multiple defect types is different, then from the defect degree information corresponding to the multiple defect types, the defect degree information with the highest defect degree information is obtained as the target defect degree information; if the defect degree information corresponding to the multiple defect types is the same, then from the defect degree information corresponding to the multiple defect types, the defect degree information of the preset defect type is obtained as the target defect degree information.
[0140] When the display screen defect has a defect type and corresponds to multiple defect degree information, the parameter acquisition module is specifically used to obtain the optimal defect compensation parameter corresponding to the defect degree information of the defect type.
[0141] Wherein, the compensation module is used to compensate the display screen defect corresponding to the defect type of the display screen according to the optimal defect compensation parameter.
[0142] In one embodiment, the at least one defect type includes color unevenness defects and / or brightness unevenness defects, and the defect features in the defect feature set corresponding to the color unevenness defect include at least one of color features and saturation features, and the defect features in the defect feature set corresponding to the brightness unevenness defect include at least one of light and dark features, horizontal and vertical stripes features, overall brightness features, local brightness features, and messy stripes features.
[0143] In one embodiment, if Fig.12 As shown, the data processing device 300 may further include a training module 305. The training module 305 is used to determine the parameters of the defect degree determination model to obtain the final defect degree determination model. The specific content of the training module 305 can be found in Fig.12 As shown in the training device of the defect degree determination model.
[0144] like Fig.12 As shown, a training device for a defect degree determination model is also provided. The training device 400 may include a training acquisition module 401 , a training feature extraction module 402 , a training model determination module 403 , a sorting prediction module 404 and an updating module 405 .
[0145] The training acquisition module 401 is used to acquire an optical image data set, wherein the optical image data set includes a plurality of training optical image data of a plurality of display screens corresponding to each defect type, the plurality of training optical image data are divided into a plurality of groups, and the plurality of training optical image data in each group correspond to defect degree sorting label information.
[0146] The training feature extraction module 402 is used to perform visual feature extraction processing on the multiple training optical image data in each group to obtain a training defect feature set corresponding to the defect type.
[0147] The training model determination module 403 is used to use the initial defect degree determination model to perform determination processing on the training defect feature set to obtain defect degree prediction information of the defect type.
[0148] The sorting prediction module 404 is used to determine the defect degree sorting prediction information corresponding to the multiple training optical image data in each group according to the defect degree prediction information.
[0149] The updating module 405 is used to update the parameters of the initial defect degree determination model according to the defect degree ranking prediction information and the defect degree ranking label information in each group, so as to obtain a defect degree determination model matching the defect type.
[0150] In one embodiment, the update module 405 is specifically used to determine the distance between the defect degree ranking prediction information and the defect degree ranking label information in each group; based on the distance, determine the similarity between the defect degree ranking prediction information and the defect degree ranking label information in each group; and use the parameters corresponding to the similarity that meets the preset conditions to update the parameters of the initial defect degree determination model to obtain a defect degree determination model that matches the defect type.
[0151] In one embodiment, the similarity includes the Spearman correlation coefficient. When the mean of the Spearman correlation coefficient in each group is the largest, it is determined that the preset condition is met, and the parameter when the mean of the Spearman correlation coefficient in each group is the largest is used as the parameter corresponding to the similarity that meets the preset condition.
[0152] During specific implementation, the above modules can be implemented as independent entities, or can be arbitrarily combined to be implemented as the same or several entities. The specific implementation of the above modules can refer to the previous method embodiments. The specific beneficial effects that can be achieved can also refer to the beneficial effects in the previous method embodiments, which will not be repeated here.
[0153] In addition, the present application also provides a computer device, such as Fig.13As shown, the computer device 500 includes a processor 501 and a memory 502. The processor 501 is electrically connected to the memory 502.
[0154] Processor 501 is the control center of computer device 500. It uses various interfaces and lines to connect various parts of the entire computer device. It executes various functions of the computer device and processes data by running or loading applications stored in memory 502 and calling data stored in memory 502, thereby monitoring the computer device as a whole.
[0155] In this embodiment, the processor 501 in the computer device 500 will load instructions corresponding to one or more application processes into the memory 502 according to the following steps, and the processor 501 will run the application / computer program stored in the memory 502 to implement various functions, such as:
[0156] Acquire optical image data of a display screen; perform visual feature extraction processing on the optical image data according to the defect type of the display screen defect to obtain a defect feature set corresponding to at least one defect type, wherein each defect type corresponds to a defect feature set; perform determination processing on the defect features in each corresponding defect feature set to obtain defect degree information corresponding to at least one defect type; obtain optimal defect compensation parameters corresponding to the defect degree information; and perform compensation processing on the display screen defect corresponding to the corresponding defect type according to the optimal defect compensation parameters.
[0157] The computer device can implement the steps in any embodiment of the data processing method provided in the embodiments of the present application, and therefore can achieve the beneficial effects that can be achieved by any data processing method provided in the embodiments of the present invention. Please refer to the previous embodiments for details and will not be repeated here.
[0158] Fig.14 The specific structural block diagram of the computer device provided in the embodiment of the present invention is shown, and the computer device can be used to implement the data processing method in any of the above embodiments. The computer device includes the following modules / units.
[0159] The RF circuit 610 is used to receive and send electromagnetic waves, realize the mutual conversion between electromagnetic waves and electrical signals, and thus communicate with a communication network or other devices. The RF circuit 610 may include various existing circuit elements for performing these functions, such as antennas, radio frequency transceivers, digital signal processors, encryption / decryption chips, user identity module (SIM) cards, memories, etc. The RF circuit 610 can communicate with various networks such as the Internet, corporate intranets, wireless networks, or communicate with other devices through wireless networks. The above-mentioned wireless networks may include cellular telephone networks, wireless local area networks, or metropolitan area networks. The above-mentioned wireless networks can use various communication standards, protocols and technologies, including but not limited to Global System for Mobile Communication (GSM), Enhanced Data GSM Environment (EDGE), Wideband Code Division Multiple Access (WCDMA), Code Division Multiple Access (CDMA), Time Division Multiple Access (TDMA), Wireless Fidelity (Wi-Fi) (such as Institute of Electrical and Electronics Engineers standards IEEE 802.11a, IEEE 802.11b, IEEE802.11g and / or IEEE802.11n), Voice over Internet Protocol (VoIP), Worldwide Interoperability for Microwave Access (Wi-Max), other protocols for email, instant messaging and short messages, and any other suitable communication protocols, even those that have not yet been developed.
[0160] The memory 620 can be used to store software programs (computer programs) and modules, such as the corresponding program instructions / modules in the above embodiments. The processor 680 executes various functional applications and data processing by running the software programs and modules stored in the memory 620. The memory 620 may include a high-speed random access memory, and may also include a non-volatile memory, such as one or more magnetic storage devices, flash memory, or other non-volatile solid-state memory. In some examples, the memory 620 may further include a memory remotely arranged relative to the processor 680, and these remote memories may be connected to the computer device 600 via a network. Examples of the above-mentioned network include, but are not limited to, the Internet, an intranet, a local area network, a mobile communication network, and combinations thereof.
[0161] The input unit 630 can be used to receive input digital or character information, and generate keyboard, mouse, joystick, optical or trackball signal input related to user settings and function control. Specifically, the input unit 630 may include a touch-sensitive surface 631 and other input devices 632. The touch-sensitive surface 631, also known as a touch display screen (touch screen) or a touchpad, can collect user touch operations on or near it (such as operations performed by users using fingers, styluses, or any other suitable objects or accessories on or near the touch-sensitive surface 631), and drive corresponding connection devices according to a pre-set program. Optionally, the touch-sensitive surface 631 may include a touch detection device and a touch controller. Among them, the touch detection device detects the user's touch orientation, detects the signal brought by the touch operation, and transmits the signal to the touch controller; the touch controller receives the touch information from the touch detection device, converts it into touch point coordinates, and then sends it to the processor 680, and can receive and execute commands sent by the processor 680. In addition, the touch-sensitive surface 631 can be implemented in various types such as resistive, capacitive, infrared, and surface acoustic wave. In addition to the touch-sensitive surface 631, the input unit 630 may further include other input devices 632. Specifically, the other input devices 632 may include, but are not limited to, one or more of a physical keyboard, function keys (such as volume control keys, switch keys, etc.), a trackball, a mouse, a joystick, and the like.
[0162] The display unit 640 can be used to display information input by the user or information provided to the user and various graphical user interfaces of the computer device 600, which can be composed of graphics, text, icons, videos and any combination thereof. The display unit 640 may include a display panel 641. Optionally, the display panel 641 may be configured in the form of LCD (Liquid Crystal Display), OLED (Organic Light-Emitting Diode), etc. Further, the touch-sensitive surface 631 may cover the display panel 641. When the touch-sensitive surface 631 detects a touch operation on or near it, it is transmitted to the processor 680 to determine the type of touch event, and then the processor 680 provides corresponding visual output on the display panel 641 according to the type of touch event. Although in the figure, the touch-sensitive surface 631 and the display panel 641 are implemented as two independent components to implement input and output functions, it can be understood that the touch-sensitive surface 631 is integrated with the display panel 641 to implement input and output functions.
[0163] The computer device 600 may also include at least one sensor 650, such as a light sensor, a direction sensor, a proximity sensor, and other sensors. As a type of motion sensor, a gravity acceleration sensor can detect the magnitude of acceleration in all directions (generally three axes), and can detect the magnitude and direction of gravity when stationary, and can be used for applications that recognize the posture of the mobile phone, vibration recognition related functions (such as pedometer, knocking), etc. As for other sensors that can be configured in the computer device 600, such as a gyroscope, a barometer, a hygrometer, a thermometer, an infrared sensor, etc., they will not be described in detail here.
[0164] The audio circuit 660, the speaker 661, and the microphone 662 can provide an audio interface between the user and the computer device 600. The audio circuit 660 can transmit the electrical signal converted from the received audio data to the speaker 661, which is converted into a sound signal for output; on the other hand, the microphone 662 converts the collected sound signal into an electrical signal, which is received by the audio circuit 660 and converted into audio data, and then the audio data is output to the processor 680 for processing, and then sent to another computer device through the RF circuit 610, or the audio data is output to the memory 620 for further processing. The audio circuit 660 may also include an earphone jack to provide communication between an external headset and the computer device 600.
[0165] The computer device 600 can help users receive requests, send information, etc. through a transmission module 670 (such as a Wi-Fi module), which provides users with wireless broadband Internet access. Although the transmission module 670 is shown in the figure, it is understandable that it is not a necessary component of the computer device 600 and can be omitted as needed without changing the essence of the invention.
[0166] The processor 680 is the control center of the computer device 600. It uses various interfaces and lines to connect various parts of the entire mobile phone. By running or executing software programs (computer programs) and / or modules stored in the memory 620, and calling data stored in the memory 620, it executes various functions of the computer device 600 and processes data, thereby monitoring the computer device as a whole. Optionally, the processor 680 may include one or more processing cores; in some embodiments, the processor 680 may integrate an application processor and a modem processor, wherein the application processor mainly processes the operating system, user interface, and application programs, etc., and the modem processor mainly processes wireless communications. It can be understood that the above-mentioned modem processor may not be integrated into the processor 680.
[0167] The computer device 600 also includes a power supply 690 (such as a battery) for supplying power to various components. In some embodiments, the power supply can be logically connected to the processor 680 through a power management system, so that the power management system can manage charging, discharging, and power consumption management. The power supply 690 can also include any components such as one or more DC or AC power supplies, recharging systems, power failure detection circuits, power converters or inverters, and power status indicators.
[0168] Although not shown, the computer device 600 also includes a camera (such as a front camera, a rear camera), a Bluetooth module, etc., which will not be described in detail here. Specifically in this embodiment, the display unit of the computer device is a touch screen display, and the computer device also includes a memory, and one or more programs (computer programs), wherein one or more programs are stored in the memory, and are configured to be executed by one or more processors to include the steps of the data processing method in any of the above embodiments.
[0169] In specific implementation, the above modules can be implemented as independent entities, or can be arbitrarily combined and implemented as the same or several entities. The specific implementation of the above modules can be found in the previous method embodiments, which will not be repeated here.
[0170] Those skilled in the art will appreciate that all or part of the steps in the various methods of the above embodiments can be completed by instructions (computer programs) or by controlling related hardware through instructions, and the instructions can be stored in a computer-readable storage medium and loaded and executed by a processor. To this end, an embodiment of the present invention provides a storage medium in which a plurality of instructions are stored, and the instructions can be loaded by a processor to execute the steps of any embodiment of the data processing method and / or the defect degree determination model training method provided by the embodiment of the present invention.
[0171] The storage medium may include: a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk, etc.
[0172] Since the instructions stored in the storage medium can execute the steps in any embodiment of the data processing method provided in the embodiments of the present invention, the beneficial effects that can be achieved by the data processing method provided in the embodiments of the present invention can be achieved. Please refer to the previous embodiments for details and will not be repeated here.
[0173] The above is a detailed introduction to a data processing method, device, storage medium and computer equipment provided in the embodiments of the present application. Specific examples are used in this article to illustrate the principles and implementation methods of the present application. The description of the above embodiments is only used to help understand the method of the present application and its core idea; at the same time, for technical personnel in this field, according to the idea of the present application, there will be changes in the specific implementation method and application scope. In summary, the content of this specification should not be understood as a limitation on the present application.
Claims
1. A data processing method, It is characterized in that include: Acquiring optical image data of the display screen; Performing visual feature extraction processing on the optical image data according to the defect type of the display screen defect to obtain a defect feature set corresponding to at least one defect type, wherein each defect type corresponds to a defect feature set; Determine and process the defect features in each corresponding defect feature set to obtain defect degree information corresponding to at least one defect type; Compensation processing is performed on display screen defects of corresponding defect types according to the defect degree information.
2. The method according to claim 1, It is characterized in that The step of compensating display screen defects of corresponding defect types according to the defect degree information comprises: Obtaining optimal defect compensation parameters corresponding to the defect degree information; The display screen defects corresponding to the corresponding defect types are compensated according to the optimal defect compensation parameters.
3. The method according to claim 2, It is characterized in that When the display screen defect has multiple defect types, the step of obtaining the optimal defect compensation parameter corresponding to the defect degree information includes: Determining target defect level information from defect level information corresponding to a plurality of defect types; Obtaining optimal defect compensation parameters corresponding to target defect degree information; When the display screen defect has one defect type, the step of obtaining the optimal defect compensation parameter corresponding to the defect degree information includes: obtaining the optimal defect compensation parameter corresponding to the defect degree information of the defect type.
4. The method according to claim 3, It is characterized in that The step of determining target defect level information from defect level information corresponding to a plurality of defect types comprises: If the defect degree information corresponding to the multiple defect types is different, the defect degree information with the highest defect degree information is obtained from the defect degree information corresponding to the multiple defect types as the target defect degree information; If the defect degree information corresponding to the multiple defect types is the same, the defect degree information of the preset defect type is obtained from the defect degree information corresponding to the multiple defect types as the target defect degree information.
5. The method according to claim 1, It is characterized in that The step of determining and processing the defect features in each corresponding defect feature set to obtain defect degree information corresponding to at least one defect type includes: Using a normalization module in a defect degree determination model that matches the defect type, the defect features in the corresponding defect feature set are normalized to obtain a normalized feature set; The degree determination module in the defect degree determination model is used to perform degree determination processing on each normalized feature in the normalized feature set to obtain defect degree information corresponding to the defect type.
6. The method according to claim 5, It is characterized in that The defect degree determination model is trained by the following steps: Acquire an optical image data set, wherein the optical image data set includes a plurality of training optical image data of a plurality of display screens corresponding to each defect type, the plurality of training optical image data are divided into a plurality of groups, and the plurality of training optical image data in each group correspond to defect degree ranking label information; Performing visual feature extraction processing on the multiple training optical image data in each group to obtain a training defect feature set corresponding to the defect type; Using the initial defect degree determination model, the training defect feature set is subjected to determination processing to obtain defect degree prediction information of the defect type; Determining defect degree ranking prediction information corresponding to the plurality of training optical image data in each group according to the defect degree prediction information; The parameters of the initial defect degree determination model are updated according to the defect degree ranking prediction information and the defect degree ranking label information in each group to obtain a defect degree determination model matching the defect type.
7. The method according to claim 6, It is characterized in that The step of updating the parameters of the initial defect degree determination model according to the defect degree ranking prediction information and the defect degree ranking label information in each group to obtain a defect degree determination model matching the defect type includes: Determining the distance between the defect level ranking prediction information and the defect level ranking label information in each group; Determining the similarity between the defect level ranking prediction information and the defect level ranking label information in each group according to the distance; The parameters of the initial defect degree determination model are updated using the parameters corresponding to the similarity that meets the preset conditions, so as to obtain a defect degree determination model that matches the defect type.
8. The method according to claim 7, It is characterized in that The similarity includes the Spearman correlation coefficient. When the mean of the Spearman correlation coefficient in each group is the largest, it is determined that the preset condition is met, and the parameter when the mean of the Spearman correlation coefficient in each group is the largest is used as the parameter corresponding to the similarity that meets the preset condition.
9. The method according to claim 1, It is characterized in that The at least one defect type includes color unevenness defects and / or brightness unevenness defects. The defect features in the defect feature set corresponding to the color unevenness defect include at least one of color features and saturation features. The defect features in the defect feature set corresponding to the brightness unevenness defect include at least one of light and dark features, horizontal and vertical stripes features, overall brightness features, local brightness features and messy stripes features.
10. A data processing device, It is characterized in that include: An image acquisition module, used to acquire optical image data of the display screen; A feature extraction module, configured to perform visual feature extraction processing on the optical image data according to the defect type of the display screen defect, so as to obtain a defect feature set corresponding to at least one defect type, wherein each defect type corresponds to a defect feature set; A determination module, used for performing determination processing on the defect features in each corresponding defect feature set to obtain defect degree information corresponding to at least one defect type; The compensation module is used to compensate display screen defects of corresponding defect types according to the defect degree information.
11. A computer-readable storage medium, It is characterized in that The computer-readable storage medium stores a plurality of instructions, and the instructions are suitable for being loaded by a processor to execute the steps in any one of the methods of claims 1 to 9.
12. A computer device, It is characterized in that It comprises a processor and a memory, wherein the processor is electrically connected to the memory, the memory is used to store instructions and data, and the processor is used to execute the steps in the method described in any one of claims 1 to 9.
Citation Information
Cited By
Defect compensation method for display panel, electronic device, and storage medium
WO2026208020A1
