OLED panel real-time digital compensation method and system

Abstract

The present application relates to the technical field of display, in particular to a kind of OLED panel real-time digital compensation method and system.A kind of OLED panel real-time digital compensation system, comprising: region segmentation module, binary static area vector construction module, hotness vector update module, compensation weight vector construction module and digital compensation module.The present application generates binary static area vector by combining the gray difference between adjacent frames, thereby identifying long-term static display area, and dynamically updates hotness vector in combination with static area map, and generates area-level compensation weight vector by fusing region brightness cumulative information, reconstructs the LUT table of OLED panel by the compensation weight vector, realizes regional, differentiated driving signal compensation, so that high-risk aging area obtains higher driving output, and then visually makes up luminance decay, significantly improves the anti-burn screen capability.

Description

Technical Field

[0001] This invention relates to the field of display technology, and specifically to a method and system for real-time digital compensation of OLED panels. Background Technology

[0002] OLED (Organic Light Emitting Diode) display technology is widely used in terminal products such as smartphones, smart wearable devices, automotive screens, and high-end televisions due to its advantages such as self-illumination, high contrast, and flexibility. However, compared with traditional LCD technology, OLED panels suffer from irreversible pixel aging during long-term use, which is particularly noticeable when displaying bright static images, making them highly susceptible to "burn-in" phenomena.

[0003] Burn-in, also known as screen burn-in, refers to the phenomenon where certain areas, due to prolonged display of fixed patterns (such as status bar icons, navigation bars, logos, etc.), cause pixels to continuously emit light and become excessively aged, resulting in problems such as decreased brightness and color shift. Even when switching to other screens, the afterimages of these patterns are still visible, severely affecting the display effect and user experience. Summary of the Invention

[0004] This invention acquires display frame images of the OLED panel in real time at a fixed sampling frequency during mobile phone use, converts them into grayscale images, and performs block processing to construct regional brightness vectors and their temporal sequences. Then, it generates binary static regional vectors by combining the grayscale differences between adjacent frames, thereby identifying long-term static display areas. The heat vector is dynamically updated in combination with the static regional map, and regional brightness accumulation information is fused to generate regional compensation weight vectors. The LUT table of the OLED panel is reconstructed through these compensation weight vectors to achieve regionalized and differentiated drive signal compensation. This allows high-risk aging areas to obtain higher drive output, thereby visually compensating for brightness decay and significantly improving screen burn-in resistance.

[0005] This invention provides a real-time digital compensation method for OLED panels, comprising:

[0006] The grayscale image of the display frame is divided into blocks to obtain several grayscale image regions of the display frame. The average grayscale values ​​of all grayscale image regions of the display frame are arranged in numerical order to form the brightness vector of the display frame image. The brightness vector of the display frame image is added to the brightness vector time sequence set corresponding to the APP called by the mobile phone at the current monitoring time point and marked with a timestamp.

[0007] The brightness vector of the display frame image corresponding to the display frame image acquired at the current monitoring time point is recorded as the target display frame image brightness vector. The brightness vector of the display frame image corresponding to the previous timestamp is selected from the brightness vector time sequence set where the target display frame image brightness vector is located and recorded as the previous display frame image brightness vector. The binary static region vector is determined based on the change of gray value between the target display frame image brightness vector and the previous display frame image brightness vector.

[0008] The heat vector is updated based on the binary static region vector, which includes the heat values ​​corresponding to all grayscale image regions of the display frames.

[0009] Add the brightness vectors of all display frame images in the time sequence set of the brightness vector where the target display frame image brightness vector is located to obtain the cumulative brightness vector, and construct the compensation weight vector based on the cumulative brightness vector and the heat vector.

[0010] The LUT table of the OLED panel is reconstructed based on the compensation weight vector. The LUT table is a mapping table that maps grayscale values ​​to the output signal of the OLED.

[0011] Preferably, the display frame grayscale image is divided into blocks to obtain several display frame grayscale image regions, specifically including the following steps:

[0012] Map the region segmentation map corresponding to the APP called by the mobile phone at the current monitoring time point to the grayscale image of the display frame. The region segmentation map includes several rectangles marked as segmentation regions. Based on the segmentation regions, several grayscale image regions of the display frame are segmented from the grayscale image of the display frame.

[0013] The region segmentation map corresponding to the APP called by the mobile phone is obtained as follows: Obtain a grayscale image of a display frame when the APP is launched by the mobile phone, and denote it as the segmented display frame grayscale image. Perform a quadtree partitioning method on the segmented display frame grayscale image, and determine whether the variance of the grayscale values ​​in each partitioned region is less than the variance threshold. If the variance of the grayscale values ​​in the partitioned region is less than the variance threshold, the partitioning stops. If the variance of the grayscale values ​​in the partitioned region is not less than the variance threshold, the quadtree partitioning method is continued to be performed on the partitioned region until the variance of the grayscale values ​​in all partitioned regions is less than the variance threshold. The smallest partitioned region is used as the unit to construct the region segmentation map, and each unit is a segmented region.

[0014] Preferably, the binary static region vector is determined based on the change in grayscale values ​​between the brightness vector of the target display frame image and the brightness vector of the preceding display frame image, specifically including the following steps:

[0015] Construct a binary static region vector. Initially, all elements of the binary static region vector are 0. Perform a difference operation on the brightness vector of the target display frame image and the brightness vector of the previous display frame image to construct a brightness difference vector of the display frame image. Determine whether the brightness difference value of the i-th display frame image in the brightness difference vector is less than the brightness change threshold. If the brightness difference value of the i-th display frame image in the brightness difference vector is less than the brightness change threshold and the number of times the brightness difference value of the i-th display frame image in the brightness difference vector is less than the brightness change threshold is higher than the duration threshold, replace the i-th value in the binary static region vector with 1, and consider it to be in a static state. Otherwise, do not operate on the binary static region vector.

[0016] Preferably, updating the heat vector based on the binary static region vector specifically includes the following steps:

[0017] In the initial state, all elements of the heat vector are 0. The heat vector is updated using the following formula: H(t+1) = αH(t) + βS, where H(t) is the heat vector before the update, H(t+1) is the heat vector after the update, α is the forgetting factor, and β is the activation factor.

[0018] Preferably, a compensation weight vector is constructed based on the brightness accumulation vector and the heat vector, specifically including the following steps:

[0019] The brightness accumulation vector, heat vector and the model of the current mobile phone OLED are concatenated to form an analysis vector. The analysis vector is then fed into the compensation weight analysis network for processing to obtain the compensation weight vector. The compensation weight analysis network is based on the BP neural network and specifically includes an input layer, a hidden layer and an output layer.

[0020] Training the compensation weight analysis network includes the following steps: obtaining several compensation weight analysis training samples, each containing analysis vectors; labeling the compensation weight analysis training samples using the compensation weight vectors; forming a compensation weight analysis training set from all labeled compensation weight analysis training samples; and training the compensation weight analysis network using the compensation weight analysis training set, with the labeled compensation weight vectors as the training target.

[0021] Preferably, reconstructing the LUT table of the OLED panel based on the compensation weight vector specifically includes the following steps:

[0022] For any grayscale image region of the display frame, the compensation weight k corresponding to the grayscale image region of the display frame is obtained from the compensation weight vector, and the grayscale image region of the display frame is mapped to the display frame image to segment the region image. A weight compensation matrix is ​​established based on the size of the region image. The weight compensation matrix stores the compensation weight k. The icon image located in the grayscale image region of the display frame at the current monitoring time point is determined by the mobile phone background. The weight compensation matrix is ​​mapped to the icon image. The values ​​corresponding to the elements in the weight compensation matrix that are not located in the icon image are replaced with 0. The elements in the weight compensation matrix correspond to the pixels in the region image, that is, the independent OLED diodes.

[0023] For any independent OLED diode, the output value in the LUT table is increased to (1+f) times the original value, where f is an element in the weight compensation matrix corresponding to the independent OLED diode.

[0024] The present invention also provides a real-time digital compensation system for OLED panels, comprising:

[0025] The region segmentation module is used to divide the grayscale image of the display frame into blocks to obtain several grayscale image regions of the display frame. The average grayscale values ​​of all grayscale image regions of the display frame are arranged in numerical order to form the brightness vector of the display frame image. The brightness vector of the display frame image is added to the brightness vector time sequence set corresponding to the APP called by the mobile phone at the current monitoring time point and marked with a timestamp.

[0026] The binary static region vector construction module is used to record the brightness vector of the display frame image corresponding to the display frame image acquired at the current monitoring time point as the target display frame image brightness vector, select the display frame image brightness vector corresponding to the previous timestamp from the brightness vector time sequence set where the target display frame image brightness vector is located and record it as the previous display frame image brightness vector, and determine the binary static region vector based on the change of gray value between the target display frame image brightness vector and the previous display frame image brightness vector.

[0027] The heat vector update module is used to update the heat vector based on the binary static region vector. The heat vector includes the heat values ​​corresponding to all grayscale image regions of the display frames.

[0028] The compensation weight vector construction module is used to perform an addition operation on all display frame image brightness vectors in the brightness vector time sequence set where the target display frame image brightness vector is located, to obtain a brightness accumulation vector, and to construct a compensation weight vector based on the brightness accumulation vector and the heat vector.

[0029] The digital compensation module is used to reconstruct the LUT table of the OLED panel based on the compensation weight vector. The LUT table is a mapping table that maps grayscale values ​​to the output signal of the OLED.

[0030] The present invention has the following advantages:

[0031] This invention acquires display frame images of the OLED panel in real time at a fixed sampling frequency during mobile phone use, converts them into grayscale images, and performs block processing to construct regional brightness vectors and their temporal sequences. Then, it generates binary static regional vectors by combining the grayscale differences between adjacent frames, thereby identifying long-term static display areas. The heat vector is dynamically updated in combination with the static regional map, and regional brightness accumulation information is fused to generate regional compensation weight vectors. The LUT table of the OLED panel is reconstructed through these compensation weight vectors to achieve regionalized and differentiated drive signal compensation. This allows high-risk aging areas to obtain higher drive output, thereby visually compensating for brightness decay and significantly improving screen burn-in resistance. Attached Figure Description

[0032] Figure 1 This is a schematic diagram of the OLED panel real-time digital compensation system used in an embodiment of the present invention. Detailed Implementation

[0033] To enable those skilled in the art to better understand the technical solutions of this invention, the technical solutions of this invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of this invention.

[0034] Example 1: A real-time digital compensation method for an OLED panel, comprising:

[0035] During mobile phone use, at monitoring time points, the display driver acquires display frame images from the OLED panel. It's important to note that the data format of these display frame images is generally RGB888, and the resolution of the display frame image matches the resolution set for the OLED panel. The time interval between adjacent monitoring time points is the same and determined by the sampling frequency, typically once every 5 frames. Then, the display frame image is converted to grayscale to obtain a grayscale image. Grayscale conversion involves converting the RGB values ​​of each pixel in the display frame image to grayscale values. The grayscale conversion method is: Y = 0.299*R + 0.587*G + 0.114*B, where Y is the grayscale value after grayscale conversion. The intensity values ​​R, G, and B represent the intensity of the red, green, and blue components in the display frame image, respectively. The display frame grayscale image is then divided into blocks to obtain several display frame grayscale image regions. The average grayscale values ​​of all display frame grayscale image regions are arranged in numerical order to form a display frame image brightness vector. This brightness vector is added to the brightness vector time series set corresponding to the APP invoked by the mobile phone at the current monitoring time point and marked with a timestamp. It should be noted that the block division method for the display frame grayscale images collected by the same APP is consistent; therefore, the obtained display frame grayscale image regions can be pre-numbered.

[0036] The brightness vector of the display frame image corresponding to the display frame image acquired at the current monitoring time point is recorded as the target display frame image brightness vector. The brightness vector of the display frame image corresponding to the previous timestamp is selected from the brightness vector time sequence set where the target display frame image brightness vector is located and recorded as the previous display frame image brightness vector. The binary static region vector is determined based on the change of gray value between the target display frame image brightness vector and the previous display frame image brightness vector. The binary static region vector stores two values, 0 and 1. If the value is 0, it means that the gray value of the corresponding region is in dynamic change. If the value is 1, it means that the gray value of the corresponding region is static. These static regions are located in fixed positions for a long time and continuously light up the same group of pixels, which makes the aging speed faster than other regions, and is prone to screen burn-in.

[0037] The heat vector is updated based on the binary static region vector. The heat vector includes the heat value corresponding to the grayscale image region of all display frames. The heat vector is initially all 0. Each called APP corresponds to a heat vector. The heat vector is in a continuous update process. The heat value represents the intensity of the aging risk accumulation of the corresponding region. The larger the heat value, the greater the probability of screen burn-in in the corresponding region.

[0038] The brightness vectors of all display frames in the time sequence set of the brightness vector of the target display frame image are added together to obtain the cumulative brightness vector. A compensation weight vector is constructed based on the cumulative brightness vector and the heat vector. The compensation weight vector here includes the compensation weights corresponding to the grayscale image areas of all display frames. The compensation weights can be used to adjust the OLED panel in the future to compensate for the grayscale image areas of display frames with faster aging speed, solve the screen burn-in problem from the perspective of viewing angle effect, and improve the user experience.

[0039] The LUT table of the OLED panel is reconstructed based on the compensation weight vector. The LUT table is a mapping table that maps grayscale values ​​to the output signals of the OLED, such as voltage, current and PWM signals. Reconstructing the LUT table can provide higher output values ​​for the grayscale image area of ​​the display frame, thereby solving the screen burn-in problem from the perspective of viewing angle.

[0040] This application acquires display frame images of the OLED panel in real time at a fixed sampling frequency during mobile phone use, converts them into grayscale images, and performs block processing to construct regional brightness vectors and their temporal sequences. Then, it generates binary static regional vectors by combining the grayscale differences between adjacent frames, thereby identifying long-term static display areas. The application also dynamically updates the heat vector by combining the static regional map and integrates the accumulated regional brightness information to generate a regional compensation weight vector. This compensation weight vector is used to reconstruct the LUT table of the OLED panel to achieve regionalized and differentiated drive signal compensation, enabling high-risk aging areas to obtain higher drive output, thereby visually compensating for brightness decay and significantly improving screen burn-in resistance.

[0041] The display frame grayscale image is divided into blocks to obtain several display frame grayscale image regions. The specific steps include the following:

[0042] Map the region segmentation map corresponding to the APP called by the mobile phone at the current monitoring time point to the grayscale image of the display frame. The region segmentation map includes several rectangles marked as segmentation regions. Based on the segmentation regions, several grayscale image regions of the display frame are segmented from the grayscale image of the display frame.

[0043] The region segmentation map corresponding to the APP called by the mobile phone is obtained as follows: A grayscale image of a display frame when the APP is launched is obtained, denoted as the segmented display frame grayscale image. A quadtree partitioning method is applied to the segmented display frame grayscale image. The quadtree partitioning method divides the region into four rectangles of the same size in each partition. It is then determined whether the variance of the grayscale values ​​in each partitioned region is less than a variance threshold (which is manually set). If the variance of the grayscale values ​​in the partitioned region is less than the variance threshold, the partitioning stops. If the variance of the grayscale values ​​in the partitioned region is not less than the variance threshold, the quadtree partitioning method is continued until the variance of the grayscale values ​​in all partitioned regions is less than the variance threshold. The smallest partitioned region is used as the unit to construct the region segmentation map, and each unit is a segmented region. It should be noted that since different APPs have different interfaces and icons, finding the smallest element using the quadtree partitioning method and using the smallest element to segment the display frame grayscale image can make subsequent brightness compensation more accurate.

[0044] The binary static region vector is determined based on the change in grayscale values ​​between the brightness vector of the target display frame image and the brightness vector of the previous display frame image. This process includes the following steps:

[0045] Construct a binary static region vector. Initially, all elements of the binary static region vector are 0. Perform a difference operation on the brightness vector of the target display frame image and the brightness vector of the previous display frame image to construct a brightness difference vector of the display frame image. Determine whether the brightness difference value of the i-th display frame image in the brightness difference vector is less than the brightness change threshold. The brightness change threshold is set manually, for example, 3.5 gray values. If the brightness difference value of the i-th display frame image in the brightness difference vector is less than the brightness change threshold and the number of times the brightness difference value of the i-th display frame image in the brightness difference vector is less than the brightness change threshold is higher than the duration threshold, the duration threshold is set manually, for example, 10. Replace the i-th value in the binary static region vector with 1, and consider it to be in a static state. Otherwise, do not operate on the binary static region vector.

[0046] It should be noted that at each timestamp in the brightness vector time sequence set, the brightness difference vector of the display frame image will be judged. At this time, a counter will be used to record the number of times the brightness difference value of the i-th display frame image in the brightness difference vector of the display frame image is less than the brightness change threshold.

[0047] The heat vector is updated based on the binary static region vector, specifically including the following steps:

[0048] In the initial state, all elements of the heat vector are 0. The heat vector is updated using the following formula: H(t+1) = αH(t) + βS, where H(t) is the heat vector before the update, H(t+1) is the heat vector after the update, α is the forgetting factor, which is generally 0.9, and β is the activation factor, which is generally 0.1. In order to improve the accuracy of brightness compensation, α and β can also be obtained through a group optimization algorithm. Based on the update formula, if the grayscale image area of ​​the display frame is continuously static, that is, the corresponding value is always 1, then the heat value will continue to increase and approach 1 through the activation factor, which can be regarded as a high-risk area for screen burn-in. Conversely, the heat value will continuously decay, which can realize the cumulative judgment of the aging degree of the grayscale image area of ​​the display frame.

[0049] The compensation weight vector is constructed based on the cumulative brightness vector and the heat vector, specifically including the following steps:

[0050] The brightness accumulation vector, heat vector and the model of the current mobile phone OLED are concatenated to form an analysis vector. The analysis vector is then fed into the compensation weight analysis network for processing to obtain the compensation weight vector. The compensation weight analysis network is based on the BP neural network and includes an input layer, a hidden layer and an output layer. It is similar to a fully connected structure and can realize the mapping between complex relationships.

[0051] The training of the compensation weight analysis network includes the following steps: Obtaining several compensation weight analysis training samples, each containing analysis vectors obtained by the operator based on actual mobile phone operation; labeling the training samples with compensation weight vectors obtained by the operator during brightness adjustment experiments; forming a compensation weight analysis training set from all labeled training samples; training the compensation weight analysis network using this training set, with the labeled compensation weight vectors as the training target; determining whether the accuracy of the compensation weight analysis network meets expectations; if the accuracy meets expectations, outputting the trained compensation weight analysis network; otherwise, continuing to train the compensation weight analysis network using the training set.

[0052] The LUT table of the OLED panel is reconstructed based on the compensation weight vector, specifically including the following steps:

[0053] For any grayscale image region of the display frame, the compensation weight k corresponding to the grayscale image region of the display frame is obtained from the compensation weight vector, and the grayscale image region of the display frame is mapped to the display frame image to segment the region image. A weight compensation matrix is ​​established based on the size of the region image. The weight compensation matrix stores the compensation weight k. The icon image located in the grayscale image region of the display frame at the current monitoring time point is determined by the mobile phone background. The weight compensation matrix is ​​mapped to the icon image. The values ​​corresponding to the elements in the weight compensation matrix that are not located in the icon image are replaced with 0. The elements in the weight compensation matrix correspond to the pixels in the region image, that is, the independent OLED diodes.

[0054] For any independent OLED diode, the output value in the LUT table is increased to (1+f) times the original value, where f is an element in the weight compensation matrix corresponding to the independent OLED diode.

[0055] Example 2, a real-time digital compensation system for an OLED panel, see [link / reference] Figure 1 ,include:

[0056] The region segmentation module is used to acquire display frame images of the OLED panel via the display driver at monitoring time points. It should be noted that the data format of the acquired display frame images is generally RGB888, and the resolution of the display frame images is consistent with the resolution set for the OLED panel. The time interval between adjacent monitoring time points is the same and determined by the sampling frequency, which is generally once every 5 frames. The display frame images are then converted to grayscale to obtain grayscale images. Grayscale conversion converts the RGB values ​​of each pixel in the display frame image to grayscale values. The grayscale conversion method is: Y = 0.299*R + 0.587*G + 0.114*B, where Y is the grayscale value after grayscale conversion. The intensity values ​​R, G, and B represent the intensity of the red, green, and blue components in the display frame image, respectively. The display frame grayscale image is then divided into blocks to obtain several display frame grayscale image regions. The average grayscale values ​​of all display frame grayscale image regions are arranged in numerical order to form a display frame image brightness vector. This brightness vector is added to the brightness vector time series set corresponding to the APP invoked by the mobile phone at the current monitoring time point and marked with a timestamp. It should be noted that the block division method for the display frame grayscale images collected by the same APP is consistent; therefore, the obtained display frame grayscale image regions can be pre-numbered.

[0057] The binary static region vector construction module is used to record the brightness vector of the display frame image corresponding to the display frame image acquired at the current monitoring time point as the target display frame image brightness vector, select the display frame image brightness vector corresponding to the previous timestamp from the brightness vector time sequence set where the target display frame image brightness vector is located and record it as the previous display frame image brightness vector, and determine the binary static region vector based on the change of gray value between the target display frame image brightness vector and the previous display frame image brightness vector. The binary static region vector stores two values, 0 and 1. If the value is 0, it means that the gray value of the corresponding region is in dynamic change. If the value is 1, it means that the gray value of the corresponding region is static. These static regions are located in fixed positions for a long time and continuously light up the same group of pixels, which causes the aging speed to be faster than other regions, and is prone to screen burn-in.

[0058] The heat vector update module is used to update the heat vector based on the binary static region vector. The heat vector includes the heat value corresponding to the grayscale image region of all display frames, and the heat vector is initially all 0. Each called APP corresponds to a heat vector, and the heat vector is in a continuous update process. The heat value represents the intensity of the aging risk accumulation of the corresponding region. The larger the heat value, the greater the probability of screen burn-in in the corresponding region.

[0059] The compensation weight vector construction module is used to perform an addition operation on the brightness vectors of all display frame images in the brightness vector time sequence set where the target display frame image brightness vector is located, to obtain a brightness accumulation vector. Based on the brightness accumulation vector and the heat vector, a compensation weight vector is constructed. The compensation weight vector here includes the compensation weights corresponding to all grayscale image areas of the display frames. The compensation weights can be used to adjust the OLED panel in the future to compensate for the grayscale image areas of display frames with faster aging speed, solve the screen burn-in problem from the perspective of viewing angle, and improve the user experience.

[0060] The digital compensation module is used to reconstruct the LUT table of the OLED panel based on the compensation weight vector. The LUT table is a mapping table that maps grayscale values ​​to the output signals of the OLED, such as voltage, current and PWM signals. By reconstructing the LUT table, higher output values ​​can be provided to the grayscale image area of ​​the display frame, thereby solving the screen burn-in problem from the perspective of viewing angle.

[0061] It should be understood that those skilled in the art can make improvements or modifications based on the above description, and all such improvements and modifications should fall within the protection scope of the appended claims. Parts not described in detail in this specification are prior art known to those skilled in the art.

Claims

1. A real-time digital compensation method for an OLED panel, characterized in that, include: The grayscale image of the display frame is divided into blocks to obtain several grayscale image regions of the display frame. The average grayscale values ​​of all grayscale image regions of the display frame are arranged in numerical order to form the brightness vector of the display frame image. The brightness vector of the display frame image is added to the brightness vector time sequence set corresponding to the APP called by the mobile phone at the current monitoring time point and marked with a timestamp. The brightness vector of the display frame image corresponding to the display frame image acquired at the current monitoring time point is recorded as the target display frame image brightness vector. The brightness vector of the display frame image corresponding to the previous timestamp is selected from the brightness vector time sequence set where the target display frame image brightness vector is located and recorded as the previous display frame image brightness vector. The binary static region vector is determined based on the change of gray value between the target display frame image brightness vector and the previous display frame image brightness vector. The heat vector is updated based on the binary static region vector, which includes the heat values ​​corresponding to all grayscale image regions of the display frames. Add the brightness vectors of all display frame images in the time sequence set of the brightness vector where the target display frame image brightness vector is located to obtain the cumulative brightness vector, and construct the compensation weight vector based on the cumulative brightness vector and the heat vector. The LUT table of the OLED panel is reconstructed based on the compensation weight vector. The LUT table is a mapping table that maps gray values ​​to the output signal of the OLED. The LUT table of the OLED panel is reconstructed based on the compensation weight vector, specifically including the following steps: For any grayscale image region of the display frame, the compensation weight k corresponding to the grayscale image region of the display frame is obtained from the compensation weight vector, and the grayscale image region of the display frame is mapped to the display frame image to segment the region image. A weight compensation matrix is ​​established based on the size of the region image. The weight compensation matrix stores the compensation weight k. The icon image located in the grayscale image region of the display frame at the current monitoring time point is determined by the mobile phone background. The weight compensation matrix is ​​mapped to the icon image. The values ​​corresponding to the elements in the weight compensation matrix that are not located in the icon image are replaced with 0. The elements in the weight compensation matrix correspond to the pixels in the region image, that is, the independent OLED diodes. For any independent OLED diode, the output value in the LUT table is increased to (1+f) times the original value, where f is an element in the weight compensation matrix corresponding to the independent OLED diode.

2. The real-time digital compensation method for an OLED panel according to claim 1, characterized in that, The display frame grayscale image is divided into blocks to obtain several display frame grayscale image regions. The specific steps include the following: Map the region segmentation map corresponding to the APP called by the mobile phone at the current monitoring time point to the grayscale image of the display frame. The region segmentation map includes several rectangles marked as segmentation regions. Based on the segmentation regions, several grayscale image regions of the display frame are segmented from the grayscale image of the display frame. The region segmentation map corresponding to the APP called by the mobile phone is obtained as follows: Obtain a grayscale image of a display frame when the APP is launched by the mobile phone, and denote it as the segmented display frame grayscale image. Perform a quadtree partitioning method on the segmented display frame grayscale image, and determine whether the variance of the grayscale values ​​in each partitioned region is less than the variance threshold. If the variance of the grayscale values ​​in the partitioned region is less than the variance threshold, the partitioning stops. If the variance of the grayscale values ​​in the partitioned region is not less than the variance threshold, the quadtree partitioning method is continued to be performed on the partitioned region until the variance of the grayscale values ​​in all partitioned regions is less than the variance threshold. The smallest partitioned region is used as the unit to construct the region segmentation map, and each unit is a segmented region.

3. The real-time digital compensation method for an OLED panel according to claim 2, characterized in that, The binary static region vector is determined based on the change in grayscale values ​​between the brightness vector of the target display frame image and the brightness vector of the previous display frame image. This process includes the following steps: Construct a binary static region vector. Initially, all elements of the binary static region vector are 0. Perform a difference operation on the brightness vector of the target display frame image and the brightness vector of the previous display frame image to construct a brightness difference vector of the display frame image. Determine whether the brightness difference value of the i-th display frame image in the brightness difference vector is less than the brightness change threshold. If the brightness difference value of the i-th display frame image in the brightness difference vector is less than the brightness change threshold and the number of times the brightness difference value of the i-th display frame image in the brightness difference vector is less than the brightness change threshold is higher than the duration threshold, replace the i-th value in the binary static region vector with 1, and consider it to be in a static state. Otherwise, do not operate on the binary static region vector.

4. The real-time digital compensation method for an OLED panel according to claim 3, characterized in that, The heat vector is updated based on the binary static region vector, specifically including the following steps: In the initial state, all elements of the heat vector are 0. The heat vector is updated using the following formula: H(t+1) = αH(t) + βS, where H(t) is the heat vector before the update, H(t+1) is the heat vector after the update, α is the forgetting factor, and β is the activation factor.

5. The real-time digital compensation method for an OLED panel according to claim 4, characterized in that, The compensation weight vector is constructed based on the cumulative brightness vector and the heat vector, specifically including the following steps: The brightness accumulation vector, heat vector and the model of the current mobile phone OLED are concatenated to form an analysis vector. The analysis vector is then fed into the compensation weight analysis network for processing to obtain the compensation weight vector. The compensation weight analysis network is based on the BP neural network and specifically includes an input layer, a hidden layer and an output layer. Training the compensation weight analysis network includes the following steps: obtaining several compensation weight analysis training samples, each containing analysis vectors; labeling the compensation weight analysis training samples using the compensation weight vectors; forming a compensation weight analysis training set from all labeled compensation weight analysis training samples; and training the compensation weight analysis network using the compensation weight analysis training set, with the labeled compensation weight vectors as the training target.

6. A real-time digital compensation system for an OLED panel, characterized in that, The system applies the real-time digital compensation method for an OLED panel according to any one of claims 1-5, including: The region segmentation module is used to divide the grayscale image of the display frame into blocks to obtain several grayscale image regions of the display frame. The average grayscale values ​​of all grayscale image regions of the display frame are arranged in numerical order to form the brightness vector of the display frame image. The brightness vector of the display frame image is added to the brightness vector time sequence set corresponding to the APP called by the mobile phone at the current monitoring time point and marked with a timestamp. The binary static region vector construction module is used to record the brightness vector of the display frame image corresponding to the display frame image acquired at the current monitoring time point as the target display frame image brightness vector, select the display frame image brightness vector corresponding to the previous timestamp from the brightness vector time sequence set where the target display frame image brightness vector is located and record it as the previous display frame image brightness vector, and determine the binary static region vector based on the change of gray value between the target display frame image brightness vector and the previous display frame image brightness vector. The heat vector update module is used to update the heat vector based on the binary static region vector. The heat vector includes the heat values ​​corresponding to all grayscale image regions of the display frames. The compensation weight vector construction module is used to perform an addition operation on all display frame image brightness vectors in the brightness vector time sequence set where the target display frame image brightness vector is located, to obtain a brightness accumulation vector, and to construct a compensation weight vector based on the brightness accumulation vector and the heat vector. The digital compensation module is used to reconstruct the LUT table of the OLED panel based on the compensation weight vector. The LUT table is a mapping table that maps grayscale values ​​to the output signal of the OLED.

Images