Method, device and electronic equipment for eliminating mura of display panel

Abstract

The application provides a method, device and electronic equipment for eliminating Mura of a display panel, the method comprising: obtaining original luminance data under at least two gray scale parameters; obtaining first compensation data according to the gray scale parameters and the original luminance data; using a luminance prediction model to obtain predicted luminance data after using the compensation data and to obtain predicted compensation data according to the predicted luminance data; and obtaining fusion compensation data according to the first compensation data and the predicted compensation data, wherein the steps of using the luminance prediction model and obtaining the predicted compensation data are repeated at least three times. The application uses the luminance prediction model to iteratively obtain the compensation data, thereby achieving multiple iterative compensations by only one shooting, effectively reducing the number of shootings and the overall processing time.

Description

Technical Field

[0001] This application relates to the field of display panel technology, and in particular to a method, apparatus and electronic device for eliminating mura on display panels. Background Technology

[0002] With the widespread application of OLED (Organic Light-Emitting Diode) display panels in the consumer electronics market, users and manufacturers are increasingly demanding higher display quality. During the manufacturing process of OLED panels, uneven brightness, known as Mura (non-uniformity), frequently occurs due to factors such as material properties, process variations, or equipment precision. Mura manifests as localized brightness differences, such as dark or bright spots, severely impacting the uniformity and visual effect of the displayed image. To improve product yield, Demura (Mura elimination) technology is widely used in the calibration stage of display panels. By adjusting the driving parameters of pixels, it compensates for brightness differences, thereby achieving overall display uniformity.

[0003] Demura technology typically uses capacitive coupling devices to capture images of the display panel at different grayscale levels, obtaining the actual brightness data of each pixel unit, and calculating compensation values ​​based on the measurement results. Traditional methods brighten overly dark areas by adjusting the grayscale value or voltage, while suppressing the output of overly bright areas, thus achieving brightness balance. This compensation method based on measured data can improve the Mura phenomenon to some extent and has become a standard processing procedure in the industry.

[0004] However, with the development of high-resolution, large-size OLED panels, the precision requirements for the Demura effect are constantly increasing. To achieve better uniformity, existing technologies often employ a multi-iterative compensation strategy, which involves repeatedly taking pictures, calculating, and compensating. While this method can gradually optimize the display effect, each iteration requires re-acquiring image data, leading to a significant increase in overall processing time. Especially in mass production environments, multiple image taking operations can drastically reduce production line cycle time and increase production costs.

[0005] Furthermore, factors such as camera stability and changes in ambient lighting during multiple photo captures can introduce measurement errors, affecting the accuracy of the compensation data. Current technology has not yet effectively resolved the balance between efficiency and accuracy; while pursuing high-quality display effects, it is difficult to avoid increased time costs. Summary of the Invention

[0006] In view of the above problems, the purpose of this application is to provide a method, apparatus and electronic device for eliminating display panel murmur, which uses a brightness prediction model to iteratively obtain compensation data, thereby enabling multiple iterative compensations to be completed with only one photo, effectively reducing the number of photos and the overall processing time.

[0007] According to one aspect of this application, a method for eliminating mura in a display panel is provided, comprising: acquiring original brightness data under at least two grayscale parameters; acquiring first compensation data based on the grayscale parameters and the original brightness data; acquiring predicted brightness data after applying the compensation data using a brightness prediction model and acquiring predicted compensation data based on the predicted brightness data; and acquiring fusion compensation data based on the first compensation data and the predicted compensation data, wherein the steps of using the brightness prediction model and acquiring the predicted compensation data are repeated at least three times.

[0008] Optionally, the training method of the brightness prediction model includes: collecting different grayscale parameters and corresponding brightness data as training data and validation data; constructing a brightness prediction model; using the grayscale data in the training data as input data of the brightness prediction model, and obtaining output data through the brightness prediction model; determining whether the error value between the output data and the corresponding brightness data is less than a threshold; determining whether all training data are less than the threshold; changing the input data of the brightness prediction model and repeating the steps of obtaining output data and comparison; when it is not less than the threshold, adjusting the brightness prediction model, and repeating the steps of obtaining output data and comparison based on the input data.

[0009] Optionally, after completing the training of the brightness prediction model, the method further includes: using the grayscale parameters in the verification data as input data for the brightness prediction model, and obtaining output data through the brightness prediction model; determining whether the error value between the output data and the brightness data is less than a threshold; determining whether all verification data are less than the threshold; changing the input data of the brightness prediction model and repeating the steps of obtaining output data and comparison; and when it is not less than the threshold, returning to the training method of the brightness prediction model.

[0010] Optionally, the same brightness data includes multiple brightness parameters, and the brightness parameters of different pixel units of the display panel are different; the brightness parameters of the same pixel unit include the brightness values ​​of the three sub-pixels R, G, and B under the same grayscale parameter.

[0011] Optionally, the steps of obtaining predicted brightness data after applying compensation data using a brightness prediction model and obtaining predicted compensation data based on the predicted brightness data include: setting a third grayscale parameter equal to the sum of the compensation data and the first grayscale parameter; inputting the first grayscale parameter to the third grayscale parameter, the first brightness data, and the second brightness data into the brightness prediction model; obtaining the predicted brightness data; obtaining the predicted compensation data based on the predicted brightness data; setting the first grayscale parameter equal to the third grayscale parameter; and repeating the above steps; wherein, in the step of setting the third grayscale parameter equal to the sum of the compensation data and the first grayscale parameter, the compensation data is the first compensation data in the first instance and the predicted compensation data thereafter.

[0012] Optionally, the predicted compensation data includes multiple data sets, and the fused compensation data is the sum of the first compensation data and the predicted compensation data.

[0013] Optionally, the brightness prediction model is a single-channel model, which can obtain the brightness data of any one of the three sub-pixels R, G, and B at a time; or the brightness prediction model is a three-channel model, which can obtain the brightness data of the three sub-pixels R, G, and B at a time.

[0014] Optionally, compensation data can be calculated using a polynomial fitting method or a gamma curve fitting method.

[0015] Optionally, the compensation data includes multiple compensation parameters, which are used to compensate for the grayscale parameters of different areas of the display panel.

[0016] According to another aspect of the present invention, an apparatus for eliminating mura in a display panel is provided, comprising: an image acquisition unit for acquiring original brightness data of the display panel under different grayscale parameters; a brightness prediction unit connected to the image acquisition unit for obtaining predicted brightness data according to a brightness prediction model; a compensation calculation unit connected to the image acquisition unit and the brightness prediction unit for obtaining multiple compensation data based on the original brightness data and the predicted brightness data; and a compensation fusion unit connected to the compensation calculation unit for fusing the multiple compensation data to obtain fused compensation data.

[0017] According to another aspect of the present invention, an electronic device is provided, which includes the above-described means for eliminating display panel murmur.

[0018] The method, apparatus, and electronic device for eliminating display panel mura provided in this application replace the time-consuming multiple camera capture steps in the traditional demura process by using a pre-trained brightness prediction model. Traditional methods require multiple iterations of "capture-compensate-recapture" to achieve high display quality, while this method only requires one capture in the initial stage. Subsequent iterations use the brightness prediction model to obtain predicted brightness data, thereby significantly reducing the time cost of the entire demura process.

[0019] Furthermore, by reducing reliance on physical camera photography, this method lowers the overall complexity of multiple Demura iterations and the operational requirements for hardware (such as CCD cameras). This makes the production process simpler and more controllable, reducing the uncertainty caused by equipment operation and waiting for photographic results.

[0020] Furthermore, through accurate prediction and compensation data calculation based on brightness prediction models (such as Gamma curve fitting), the brightness uniformity of the adjusted display panel can be ensured, thereby ultimately improving the display quality of the OLED display panel and meeting the market's requirements for high-performance panels. Attached Figure Description

[0021] The above and other objects, features and advantages of this application will become clearer from the following description of embodiments with reference to the accompanying drawings, in which:

[0022] Figure 1 A flowchart illustrating a method for eliminating display panel murmurs according to an embodiment of this application is shown;

[0023] Figure 2 A sub-flowchart of step S30 in a method for eliminating display panel murmur according to an embodiment of this application is shown;

[0024] Figure 3 A flowchart illustrating the training of a brightness prediction model in a method for eliminating display panel murmurs according to an embodiment of this application is shown;

[0025] Figure 4 A training structure diagram of the brightness prediction model in the method for eliminating display panel murmur according to an embodiment of this application is shown;

[0026] Figure 5 A structural diagram of an apparatus for eliminating display panel murmur according to an embodiment of this application is shown. Detailed Implementation

[0027] The present application will now be described in more detail with reference to the accompanying drawings. In the various drawings, the same elements are indicated by similar reference numerals. For clarity, the various parts in the drawings are not drawn to scale. Furthermore, some well-known parts may not be shown. For simplicity, the semiconductor structure obtained after several steps can be depicted in a single figure.

[0028] It should be understood that when describing the structure of a device, when referring to a layer or region as being "above" or "on top of" another layer or region, it can mean that it is directly above another layer or region, or that there are other layers or regions between it and another layer or region. Furthermore, if the device is flipped, that layer or region will be located "below" or "under" another layer or region.

[0029] To describe a situation where it is directly above another layer or another area, this article will use expressions such as "directly above" or "above and adjacent to".

[0030] Many specific details of this application, such as the structure, materials, dimensions, processing techniques, and methods of the devices, are described below to provide a clearer understanding of the application. However, as those skilled in the art will understand, this application may be implemented without adhering to these specific details.

[0031] The specific implementation methods of this application will be further described in detail below with reference to the accompanying drawings and embodiments.

[0032] Figure 1 A flowchart illustrating a method for eliminating display panel murmurs according to an embodiment of this application is shown; Figure 2 A sub-flowchart of step S30 in a method for eliminating display panel murmur according to an embodiment of this application is shown; Figure 3 A flowchart illustrating the training of a brightness prediction model in a method for eliminating display panel murmurs according to an embodiment of this application is shown; Figure 4 A training structure diagram of the brightness prediction model in the method for eliminating display panel murmur according to an embodiment of this application is shown.

[0033] like Figures 1 to 4 As shown, the method for eliminating display panel murmurs according to this application includes the following steps.

[0034] Step S10: Obtain the original brightness data under at least two grayscale parameters.

[0035] In this step, the image acquisition unit takes a picture of the display panel to obtain the raw brightness data of the display panel under different grayscale parameters.

[0036] The process involves providing a grayscale parameter (Gray) to the manufactured display panel, and then obtaining the raw brightness data (Rawdata) of the display panel under that grayscale parameter (Gray). The grayscale parameter (Gray) provided to the display panel is then changed, and the raw brightness data (Rawdata) of the display panel under that grayscale parameter (Gray) is obtained again. This process is repeated multiple times to obtain multiple grayscale parameters (Gray) and their corresponding raw brightness data (Rawdata).

[0037] During the manufacturing process of display panels, factors such as material properties, process variations, or equipment precision can lead to uneven brightness. This results in different areas of the same display panel having different brightness parameters when the same grayscale parameter (Gray) is applied. More specifically, under the same grayscale parameter (Gray), different pixel units within the display panel will have different brightness parameters.

[0038] In addition, since a pixel unit of the display panel has three sub-pixels R, G, and B, each sub-pixel has a brightness value.

[0039] Therefore, the raw brightness data obtained in this step includes multiple brightness parameters. The brightness parameters differ in different areas of the display panel, or in other words, the brightness parameters differ in different pixel units of the display panel. The brightness parameters of the same pixel unit include the brightness values ​​of the three sub-pixels R, G, and B under the same grayscale parameter. All subsequent raw brightness data will include multiple brightness parameters, and each brightness parameter will include three brightness values.

[0040] In some embodiments, the image acquisition unit is, for example, a CCD industrial camera, which acquires raw brightness data (Rawdata) of the display panel under different grayscale parameters by taking pictures.

[0041] Step S20: Obtain first compensation data based on the grayscale parameters and the original brightness data.

[0042] In this step, the compensation calculation unit calculates the first compensation data offset1 based on the grayscale parameter Gray and the corresponding brightness data Rawdata provided by the image acquisition unit.

[0043] In this embodiment, the compensation calculation unit calculates the compensation data offset according to the polynomial fitting method or the gamma curve fitting method.

[0044] In this embodiment, the compensation data offset includes multiple compensation parameters. The brightness compensation parameters are different for different areas of the display panel, or in other words, the compensation parameters are different for different pixel units of the display panel. The compensation parameters for the same pixel unit include the compensation values ​​of the three sub-pixels R, G, and B under the same grayscale parameter. All subsequent compensation data offsets include multiple compensation parameters, and each compensation parameter includes three compensation values.

[0045] Specifically, the formulas and steps used in calculating the compensation value of the G sub-pixel using the gamma curve fitting method include:

[0046] (1-1)

[0047] Formula (1-1) represents the relationship between the grayscale parameter Gray and the brightness data of the display panel. In this formula, L... v This represents the brightness data of the display panel; the image acquisition unit acquires L. v Rawdata is then obtained; K is a linear parameter related to the brightness data; Gray represents the grayscale parameter of the display panel, used to describe the light emission characteristics of the pixel units in the display panel; gamma represents the exponential parameter of the grayscale parameter.

[0048] According to formula (1-1), when the target brightness data of the G sub-pixel is known, the following formula (1-2) can be obtained.

[0049] (1-2)

[0050] In formula (1-2), Gray in This represents the input value for the grayscale parameter; Gray out Indicates the output value of the grayscale parameter; K C and gamma C The two parameters are used to describe the target brightness data, that is, the brightness data to be achieved after compensation. Among them, K C Specifically, it is the average value calculated based on multiple pixel units in the central area of ​​the display panel; K t and gamma t These are compensation parameters used to describe sub-pixels.

[0051] According to formula (1-2), the grayscale parameter Gray of the compensated G sub-pixel is... out It satisfies the following formula (1-3).

[0052] (1-3)

[0053] In formula (1-3), Gray inThis represents the input value for the grayscale parameter; Gray out Indicates the output value of the grayscale parameter; K C and gamma C The two parameters are used to describe the target brightness data, that is, the brightness data to be achieved after compensation; K t and gamma t These are compensation parameters used to describe sub-pixels.

[0054] According to formula (1-3), the output value of the grayscale parameter of the G sub-pixel can be calculated, that is, the grayscale value corresponding to the G sub-pixel when it is at the target brightness value. Therefore, the compensation value of the G sub-pixel is:

[0055] (1-4)

[0056] In formula (1-4), Gray in This represents the input value for the grayscale parameter; Gray out This represents the output value of the grayscale parameter, Gray. offset This represents the compensation value for the G sub-pixel.

[0057] Similarly, the calculation process for the compensation values ​​of the other two R and B sub-pixels in the same pixel unit also uses the same method.

[0058] In this embodiment, the compensation data offset is a set of grayscale compensation values ​​for each sub-pixel.

[0059] Step S30: Use the brightness prediction model to obtain the predicted brightness data after using the compensation data, and obtain the predicted compensation data based on the predicted brightness data.

[0060] In this embodiment, at least two grayscale parameters Gray and corresponding raw brightness data Rawdata and first compensation data are provided to the brightness prediction unit, so that the brightness prediction model in the brightness prediction unit obtains the predicted brightness data Rawdata.

[0061] Furthermore, it also includes: the compensation calculation unit calculates the predicted compensation data Rawdata based on the predicted brightness data Rawdata output by the brightness prediction unit.

[0062] The steps of repeating the brightness prediction model and obtaining prediction compensation data are repeated at least twice to obtain multiple prediction compensation data (Rawdata).

[0063] Specifically, this step will be described in detail using the G sub-pixel as an example, such as... Figure 2As shown. Let the grayscale parameters of the G sub-pixel obtained in step S10 be Gray1 and Gray2, the corresponding original brightness data be Rawdata1 and Rawdata2, and the first compensation data calculated in step S20 be offset1.

[0064] Step S31: Set the third grayscale parameter to be equal to the sum of the compensation data and the first grayscale parameter.

[0065] In this step, the third grayscale parameter Gray3 is set to be equal to the sum of the compensation data offset1 and the first grayscale parameter Gray1, i.e., Gray3 = Gray1 + offset1.

[0066] Step S32: Input the first grayscale parameters to the third grayscale parameters, the first brightness data, and the second brightness data into the brightness prediction model.

[0067] In this step, the brightness prediction model receives input values ​​Gray1, Gray2, Gray3, Rawdata1, and Rawdata2.

[0068] Step S33: Obtain predicted brightness data.

[0069] In this step, the brightness prediction model directly outputs the predicted brightness data Rawdata3 based on the input values ​​Gray1, Gray2, Gray3, Rawdata1, and Rawdata2.

[0070] Step S34: Obtain prediction compensation data based on the predicted brightness data.

[0071] In this step, the compensation calculation unit calculates the predicted compensation data offset2 based on the predicted brightness data Rawdata3 and the third grayscale parameter Gray3.

[0072] Step S35: Set the first grayscale parameter to equal the third grayscale parameter.

[0073] In this step, let the first grayscale parameter Gray1 equal the third grayscale parameter Gray3. Then, the new first grayscale parameter Gray1 is equal to the sum of Gray1 and offset1 in step S31. The purpose of this step is to facilitate subsequent loop steps.

[0074] Step S36: Repeat steps S31-S35 above.

[0075] In this embodiment, the brightness prediction model determines the number of repetitions of steps S31 to S34 based on the predicted brightness data.

[0076] In step S31, when this step is executed for the first time, the compensation data is the first compensation data calculated in step S20. In subsequent steps, the compensation data is the predicted compensation data calculated in the previous step of the current step.

[0077] To facilitate the description of subsequent steps, the numbers of the third grayscale parameter, predicted brightness data, and predicted compensation data in the loop are sequentially increased to make it easier to distinguish different loop steps.

[0078] In an embodiment that repeats steps S31-S34, the first step S31-S35 includes a third grayscale parameter Gray3, predicted brightness data Rawdata3, and predicted compensation data offset2; in the second step S31-S35, the third grayscale parameter Gray4 = Gray3 + offset2, and the predicted brightness data Rawdata4 and predicted compensation data offset3 can be obtained based on the third grayscale parameter Gray4; similarly, in the third step S31-S35, the third grayscale parameter Gray5 = Gray4 + offset3, and the predicted brightness data Rawdata5 and predicted compensation data offset4 can be obtained based on the third grayscale parameter Gray5; the above cycle is repeated until the number of cycles is reached or the target brightness data is reached.

[0079] In some embodiments, steps S31-S34 are repeated, for example, 3-5 times. This is because if the number of repetitions is less than 3, the final compensation effect may not reach the expected goal, while if the number of repetitions is more than 5, the usage time will be longer.

[0080] In other embodiments, the number of repetitions of steps S31-S34 can be set as needed. It is understood that the more times steps S31-S34 are repeated, the better the final compensation effect will be.

[0081] Since steps S31-35 are repeated multiple times, multiple prediction compensation data can be obtained.

[0082] In some embodiments, the brightness prediction model is a single-channel model, which can obtain the predicted brightness data of any one of the three sub-pixels R, G, and B at a time.

[0083] In other embodiments, the brightness prediction model is a three-channel model, which can obtain the predicted brightness data of three sub-pixels (R, G, and B) at one time.

[0084] Furthermore, such as Figure 3 As shown, the training steps for the brightness prediction model include:

[0085] Step S101: Collect different grayscale parameters and corresponding brightness data as training data and validation data.

[0086] In this step, a large amount of grayscale parameters (Gray) and corresponding brightness data (Rawdata) are collected from the display panel manufacturing plant. The collected grayscale parameters (Gray) and corresponding brightness data (Rawdata) are divided into training data and validation data.

[0087] Training data, for example, accounts for 80% of the total data and is used to train the brightness prediction model; validation data, for example, accounts for 20% of the total data and is used to validate the accuracy of the brightness prediction model.

[0088] In addition, in a set of grayscale parameters Gray and corresponding brightness data Rawdata, the grayscale parameters Gray are used as input data and the corresponding brightness data Rawdata are used as training labels. That is, the output data of the brightness prediction model should be the same as the brightness data Rawdata.

[0089] Step S102: Construct a brightness prediction model.

[0090] In this step, the neural network layers in the brightness prediction model are constructed, such as... Figure 4 As shown.

[0091] The brightness prediction model includes at least an input layer, a feedback layer, and an output layer. The input layer takes in compensation data, grayscale parameters (Gray), and brightness data (Rawdata), and includes at least five input terminals: one for compensation data, two for grayscale parameters (Gray), and two for brightness data (Rawdata). The feedback layer feeds back the results from subsequent layers to the preceding layers, introducing dynamic adjustment to make the final output more accurate. Furthermore, the final output can be adjusted by modifying factors within the feedback layer. The output layer outputs the predicted brightness data.

[0092] Step S103: Use the grayscale data in the training data as the input data of the brightness prediction model, and obtain the output data through the brightness prediction model.

[0093] In this step, the grayscale data (Gray) from the training data is input into the brightness prediction model, and the brightness prediction model obtains an output data, which is the predicted brightness data.

[0094] Step S104: Determine whether the error value between the output data and the corresponding brightness data is less than a threshold.

[0095] In this step, the output data of the brightness prediction model is compared with the actual brightness data corresponding to the grayscale data of the output data to determine whether the error between the output data and the actual brightness data is less than the threshold.

[0096] If the error value is less than the threshold, proceed to step S105; if the error value is not less than the threshold, proceed to step S106.

[0097] Step S105: Determine whether all training data are less than the threshold.

[0098] In this step, when the value is less than the threshold, it is further determined whether all grayscale parameters in the training data have been input into the brightness prediction model, and whether the error value between the obtained output data and the actual brightness data is less than the threshold.

[0099] Step S106: Replace the input data of the brightness prediction model and repeat the steps of acquiring output data and comparison.

[0100] In this step, if the error between the output data and the actual brightness data is less than the threshold, it means that the brightness prediction model is correct at least in this training.

[0101] The steps S103 and S104 are repeated to change the input data of the brightness prediction model, that is, to re-input the gray level parameter Gray from the training data to the brightness prediction model, and to obtain the output data of the brightness prediction model again and compare the output data with the actual brightness data, until the error value of all gray level parameters in the training data between the output data obtained by the brightness prediction model and the actual brightness data is less than the threshold.

[0102] Step S107: When the brightness is not less than the threshold, adjust the brightness prediction model and repeat the steps of obtaining output data and comparison based on the input data.

[0103] In this step, if the error between the output data and the actual brightness data is not less than the threshold, it means that the prediction error of the brightness prediction model is large, at least in this prediction.

[0104] Replace the factors in the feedback layer of the brightness prediction model to reduce the error between the output data of the brightness prediction model and the actual brightness data. Then re-input the grayscale parameter into the brightness prediction model and repeat steps S103 and S104.

[0105] Step S108: Use the grayscale parameters in the verification data as input data for the brightness prediction model, and obtain output data through the brightness prediction model.

[0106] In this step, the brightness prediction model is trained using training data. At this point, the error between the output data and the actual brightness data obtained after inputting the grayscale parameters from the training data into the brightness prediction model is less than the threshold.

[0107] Furthermore, the grayscale parameters from the validation data are input into the brightness prediction model to obtain the output data.

[0108] Step S109: Determine whether the error value between the output data and the brightness data is less than a threshold.

[0109] In this step, the output data of the brightness prediction model is compared with the actual brightness data corresponding to the grayscale data of the output data to determine whether the error between the output data and the actual brightness data is less than the threshold.

[0110] If the error value is less than the threshold, proceed to step S110; if the error value is not less than the threshold, proceed to step S111.

[0111] Step S110: Determine that all verification data are less than the threshold.

[0112] In this step, when the value is less than the threshold, it is further determined whether all grayscale parameters in the verification data have been input into the brightness prediction model, and whether the error value between the obtained output data and the actual brightness data is less than the threshold.

[0113] Step S111: Change the input data of the brightness prediction model and repeat the steps of acquiring output data and comparison.

[0114] In this step, if the error between the output data and the actual brightness data is less than the threshold, it means that the brightness prediction model is correct at least in this verification.

[0115] The steps S108 and S109 involve changing the input data of the brightness prediction model, i.e., re-inputting the grayscale parameter Gray from the verification data into the brightness prediction model, obtaining the output data of the brightness prediction model again, and comparing the output data with the actual brightness data, until the error values ​​of all grayscale parameters in the verification data between the output data obtained through the brightness prediction model and the actual brightness data are all less than the threshold.

[0116] Step S112: When the value is not less than the threshold, return to the training method of the brightness prediction model.

[0117] In this step, if the error between the output data and the actual brightness data is not less than the threshold, it means that at least in this verification, the prediction error of the brightness prediction model is large, and the process returns to step S103.

[0118] In some embodiments, the training step may be returned only if the error between the output data corresponding to a certain proportion of grayscale parameters in the verification data and the actual brightness data is not less than a threshold.

[0119] Step S40: Obtain fused compensation data based on the first compensation data and the predicted compensation data.

[0120] In this step, the compensation fusion unit fuses the first compensation data obtained in step S20 and the multiple predicted compensation data obtained in step S30 to obtain fused compensation data.

[0121] In some embodiments, the fused compensation data is, for example, the sum of a first compensation data and multiple predicted compensation data.

[0122] Figure 5 A structural diagram of an apparatus for eliminating display panel murmur according to an embodiment of this application is shown.

[0123] like Figure 5 As shown, the apparatus for eliminating display panel murmurs is the same as the aforementioned method for eliminating display panel murmurs, and includes: an image acquisition unit 210 for acquiring original brightness data of the display panel under different grayscale parameters; a brightness prediction unit 220 connected to the image acquisition unit 210 for obtaining predicted brightness data according to a brightness prediction model; a compensation calculation unit 230 connected to the image acquisition unit 210 and the brightness prediction unit 220 for obtaining multiple compensation data based on the original brightness data and the predicted brightness data; and a compensation fusion unit 240 connected to the compensation calculation unit 230 for fusing the multiple compensation data to obtain fused compensation data.

[0124] Furthermore, this application also provides an electronic device including the means for eliminating display panel murmur as described above.

[0125] The method, apparatus, and electronic device for eliminating display panel mura provided in this application replace the time-consuming multiple camera capture steps in the traditional demura process by using a pre-trained brightness prediction model. Traditional methods require multiple iterations of "capture-compensate-recapture" to achieve high display quality, while this method only requires one capture in the initial stage. Subsequent iterations use the brightness prediction model to obtain predicted brightness data, thereby significantly reducing the time cost of the entire demura process.

[0126] Furthermore, by reducing reliance on physical camera photography, this method lowers the overall complexity of multiple Demura iterations and the operational requirements for hardware (such as CCD cameras). This makes the production process simpler and more controllable, reducing the uncertainty caused by equipment operation and waiting for photographic results.

[0127] Furthermore, through accurate prediction and compensation data calculation based on brightness prediction models (such as Gamma curve fitting), the brightness uniformity of the adjusted display panel can be ensured, thereby ultimately improving the display quality of the OLED display panel and meeting the market's requirements for high-performance panels.

[0128] As described above, these embodiments of the present application do not exhaustively cover all details, nor do they limit the invention to the specific embodiments described. Clearly, many modifications and variations can be made based on the above description. This specification selects and specifically describes these embodiments to better explain the principles and practical applications of the present application, thereby enabling those skilled in the art to make good use of the present application and modifications based on it. The present application is limited only by the claims and their full scope and equivalents.

Claims

1. A method for eliminating murmur on a display panel, wherein, include: Obtain raw brightness data for at least two grayscale parameters; The first compensation data is obtained based on the grayscale parameters and the original brightness data; Use a brightness prediction model to obtain predicted brightness data after using compensation data, and obtain predicted compensation data based on the predicted brightness data; The fused compensation data is obtained based on the first compensation data and the predicted compensation data. The steps of repeating the brightness prediction model and obtaining prediction compensation data are performed at least three times.

2. The method according to claim 1, wherein, The training method for the brightness prediction model includes: Collect different grayscale parameters and corresponding brightness data as training and validation data; Construct a brightness prediction model; The grayscale data in the training data is used as the input data of the brightness prediction model, and the output data is obtained through the brightness prediction model. Determine whether the error value between the output data and the corresponding brightness data is less than a threshold; Determine if all training data are less than a threshold; Replace the input data of the brightness prediction model and repeat the steps of acquiring output data and comparison; When the brightness is not less than the threshold, the brightness prediction model is adjusted, and the steps of obtaining output data and comparison based on the input data are repeated.

3. The method according to claim 2, wherein, After training the brightness prediction model, the following steps are also included: The grayscale parameters in the verification data are used as input data for the brightness prediction model, and the output data is obtained through the brightness prediction model. Determine whether the error value between the output data and the brightness data is less than a threshold; Determine whether all verification data are less than the threshold. Replace the input data of the brightness prediction model and repeat the steps of acquiring output data and comparison; If the value is not less than the threshold, return to the training method of the brightness prediction model.

4. The method according to any one of claims 1-3, wherein, The same brightness data includes multiple brightness parameters, and the brightness parameters of different pixel units on the display panel are different; The brightness parameters of the same pixel unit include the brightness values ​​of the three sub-pixels R, G, and B under the same grayscale parameters.

5. The method according to claim 4, wherein, The steps of obtaining predicted brightness data after applying compensation data using a brightness prediction model and obtaining predicted compensation data based on the predicted brightness data include: Set the third grayscale parameter to be equal to the sum of the compensation data and the first grayscale parameter; Input the first grayscale parameters to the third grayscale parameters, the first brightness data, and the second brightness data into the brightness prediction model; Obtain predicted brightness data; Predictive compensation data is obtained based on the predicted brightness data; Set the first grayscale parameter to equal the third grayscale parameter; Repeat the above steps; In the step of setting the third grayscale parameter to be equal to the sum of the compensation data and the first grayscale parameter, the compensation data is the first compensation data in the first instance and the predicted compensation data in subsequent instances.

6. The method according to claim 5, wherein, The predicted compensation data includes multiple data sets, and the fused compensation data is the sum of the first compensation data and the predicted compensation data.

7. The method according to claim 5, wherein, The brightness prediction model is a single-channel model, which can obtain the brightness data of any one of the three sub-pixels (R, G, B) at a time; or The brightness prediction model is a three-channel model, which can obtain the brightness data of three sub-pixels (R, G, and B) at one time.

8. The method according to claim 1, wherein, Compensation data is obtained by calculating using polynomial fitting or gamma curve fitting methods.

9. The method according to claim 8, wherein, The compensation data includes multiple compensation parameters, which are used to compensate for the grayscale parameters of different areas of the display panel.

10. An apparatus for eliminating display panel murmur, wherein, include: The image acquisition unit is used to acquire the original brightness data of the display panel under different grayscale parameters; A brightness prediction unit is connected to the image acquisition unit and obtains predicted brightness data based on a brightness prediction model. A compensation calculation unit, connected to the image acquisition unit and the brightness prediction unit, is used to obtain multiple compensation data based on the original brightness data and the predicted brightness data; The compensation fusion unit, connected to the compensation calculation unit, is used to fuse multiple compensation data to obtain fused compensation data.

11. An electronic device, wherein, Includes the device for eliminating display panel murmur as described in claim 10.

Images