Adjustment method and device of display panel, electronic equipment and storage medium

Abstract

The disclosure provides an adjusting method and device of a display panel, electronic equipment and a storage medium, and relates to the technical field of display. The specific implementation scheme is as follows: a first gamma parameter of a display panel under a first refresh frequency is obtained, the first gamma parameter comprises a corresponding relationship between a gray scale and a brightness under a plurality of color modes and in conformity with a standard gamma curve; the first gamma parameter is input into a prediction model to obtain a second gamma parameter of the display panel under a second refresh frequency, wherein the prediction model comprises a down-sampling network structure and an up-sampling network structure, and the second refresh frequency is different from the first refresh frequency; and the display panel is adjusted according to the second gamma parameter. According to the technical scheme of the disclosure, the Gamma curve adjustment time of a single frequency mode screen can be consumed only, the Gamma curve adjustment of a multi-refresh frequency screen is realized, the adjustment time is saved by several times, the screen production efficiency is improved, and the cost is saved.

Description

Technical Field

[0001] This disclosure relates to the field of display technology, and more particularly to a method, apparatus, electronic device, and storage medium for adjusting a display panel. Background Technology

[0002] In the display industry, the Gamma 2.2 curve is the recognized industry standard. Therefore, Gamma curve adjustment is essential during the display screen manufacturing process. With the rapid development of display technology and the continuous upgrading of consumer demands, screens now support multiple frequency modes, and this trend continues. However, the more screen frequencies a screen has, the more Gamma values ​​need to be adjusted. Currently, the method for adjusting the Gamma curve of multi-refresh-rate screens is the same as that of single-refresh-rate screens: for each frequency mode, binding points need to be set and adjusted individually, which is time-consuming and labor-intensive, directly increasing the cost of screen production. Summary of the Invention

[0003] This disclosure provides a method, apparatus, electronic device, and storage medium for adjusting a display panel.

[0004] According to a first aspect of this disclosure, a method for adjusting a display panel is provided, comprising:

[0005] Obtain the first gamma parameter of the display panel at the first refresh rate. The first gamma parameter includes the correspondence between grayscale and brightness in multiple color modes that conform to the standard gamma curve.

[0006] The first gamma parameter is input into the prediction model to obtain the second gamma parameter of the display panel at the second refresh frequency. The prediction model includes a downsampling network structure and an upsampling network structure. The second refresh frequency is different from the first refresh frequency.

[0007] Adjust the display panel according to the second gamma parameter.

[0008] In one implementation, the first gamma parameter is input into the prediction model to obtain the second gamma parameter of the display panel at the second refresh rate, including:

[0009] The first gamma parameter is subjected to N sets of downsampling operations in an iterative manner through a downsampling network structure to obtain the target downsampling result, where N is a positive integer, and each set of downsampling operations includes convolution and pooling operations;

[0010] The upsampling network structure iteratively performs N sets of upsampling operations on the target downsampling result to obtain the target upsampling result. Each set of upsampling operations includes a deconvolution operation and a feature fusion operation. The feature map corresponding to the deconvolution result obtained by the deconvolution operation is the same size as the feature map corresponding to the convolution result obtained by the corresponding set of downsampling operations. The feature fusion operation is used to fuse the features of the deconvolution result obtained by the deconvolution operation and the convolution result obtained by the corresponding set of downsampling operations.

[0011] The second gamma parameter is obtained based on the target upsampling result.

[0012] In one embodiment, the upsampling network structure includes a first upsampling network structure and a second upsampling network structure. The first gamma parameter is input into the prediction model to obtain the second gamma parameter of the display panel at the second refresh rate, including:

[0013] The first gamma parameter is subjected to M sets of downsampling operations in an iterative manner through a downsampling network structure to obtain the target downsampling result, where M is an integer greater than 1, and each set of downsampling operations includes convolution and pooling operations.

[0014] The first upsampling network structure iteratively performs M-1 sets of first upsampling operations on the target downsampling result to obtain the first upsampling result. Each set of first upsampling operations includes a deconvolution operation and a feature fusion operation. The feature map corresponding to the deconvolution result obtained by the deconvolution operation is the same size as the feature map corresponding to the convolution result obtained by the corresponding set of downsampling operations. The feature fusion operation is used to fuse the features of the deconvolution result obtained by the deconvolution operation and the convolution result obtained by the corresponding set of downsampling operations.

[0015] The second upsampling network structure performs a second upsampling operation on the first upsampling result to obtain a second upsampling result. The second upsampling operation includes an upsampling operation and a convolution operation.

[0016] The second gamma parameter is obtained based on the second upsampling result.

[0017] In one implementation, the prediction model further includes a convolutional attention module, which inputs the first gamma parameter into the prediction model to obtain the second gamma parameter of the display panel at the second refresh rate, including:

[0018] The first downsampling result is obtained by iteratively performing N sets of downsampling operations on the first gamma parameter through a downsampling network structure, where N is a positive integer and each set of downsampling operations includes convolution and pooling operations.

[0019] The first downsampling result is processed by a convolutional attention module to obtain the second downsampling result;

[0020] The upsampling network structure iteratively performs N sets of upsampling operations on the second downsampling result to obtain the target upsampling result. Each set of upsampling operations includes deconvolution and feature fusion operations. The feature map corresponding to the deconvolution result obtained by the deconvolution operation is the same size as the feature map corresponding to the convolution result obtained by the corresponding set of downsampling operations. The feature fusion operation is used to fuse the features of the deconvolution result obtained by the deconvolution operation and the convolution result obtained by the corresponding set of downsampling operations.

[0021] The second gamma parameter is obtained based on the target upsampling result.

[0022] In one implementation, inputting the first gamma parameter into the prediction model includes:

[0023] The first gamma parameter is input into the prediction model in the form of a1×b1×c1, where a1 is the number of brightness levels of the display panel, b1 is the number of gray levels of the display panel, and c1 is the number of color modes of the display panel.

[0024] In one implementation, the sampling result obtained by the downsampling operation or the upsampling operation includes a×b×c feature data, where a, b, and c are all positive integers, a represents the height of the feature map obtained by performing the downsampling operation or the upsampling operation, b represents the width of the feature map obtained by performing the downsampling operation or the upsampling operation, and c represents the number of channels of the feature map obtained by performing the downsampling operation or the upsampling operation.

[0025] According to a second aspect of this disclosure, an adjustment device for a display panel is provided, comprising:

[0026] The acquisition module is used to acquire the first gamma parameter of the display panel at the first refresh rate. The first gamma parameter includes the correspondence between grayscale and brightness in multiple color modes that conform to the standard gamma curve.

[0027] The input module is used to input the first gamma parameter into the prediction model to obtain the second gamma parameter of the display panel at the second refresh frequency. The prediction model includes a downsampling network structure and an upsampling network structure. The second refresh frequency is different from the first refresh frequency.

[0028] The adjustment module is used to adjust the display panel according to the second gamma parameter.

[0029] In one implementation, the input module includes:

[0030] The downsampling unit is used to perform N sets of downsampling operations on the first gamma parameter in an iterative manner through the downsampling network structure to obtain the target downsampling result, where N is a positive integer, and each set of downsampling operations includes convolution and pooling operations;

[0031] The upsampling unit is used to iteratively perform N sets of upsampling operations on the target downsampling result through the upsampling network structure to obtain the target upsampling result. Each set of upsampling operations includes a deconvolution operation and a feature fusion operation. The feature map corresponding to the deconvolution result obtained by the deconvolution operation is the same size as the feature map corresponding to the convolution result obtained by the corresponding set of downsampling operations. The feature fusion operation is used to fuse the features of the deconvolution result obtained by the deconvolution operation and the convolution result obtained by the corresponding set of downsampling operations.

[0032] The parameter acquisition unit is used to obtain the second gamma parameter based on the target upsampling result.

[0033] According to a third aspect of this disclosure, an electronic device is provided, comprising:

[0034] At least one processor; and

[0035] The memory is communicatively connected to the at least one processor; wherein,

[0036] The memory stores instructions that can be executed by the at least one processor to enable the at least one processor to perform the method in the first aspect described above.

[0037] According to a fourth aspect of this disclosure, a non-transient computer-readable storage medium is provided having a computer program stored thereon that, when executed by a processor, implements the method of the first aspect described above.

[0038] According to the technology disclosed herein, the Gamma curve adjustment of a multi-refresh-rate screen can be achieved by only consuming the Gamma curve adjustment time of a single-frequency mode screen, which saves adjustment time by a factor of two, improves screen production efficiency, and saves costs.

[0039] It should be understood that the description in this section is not intended to identify key or essential features of the embodiments of this disclosure, nor is it intended to limit the scope of this disclosure. Other features of this disclosure will become readily apparent from the following description. Attached Figure Description

[0040] The accompanying drawings are provided to better understand this solution and do not constitute a limitation of this disclosure. Wherein:

[0041] Figure 1 This is a schematic flowchart of a display panel adjustment method according to an embodiment of the present disclosure;

[0042] Figure 2 This is a schematic diagram of the structure of a prediction model according to an embodiment of the present disclosure. Figure 1 ;

[0043] Figure 3 This is a schematic diagram of the structure of a prediction model according to an embodiment of the present disclosure. Figure 2 ;

[0044] Figure 4 This is a schematic diagram of the structure of a prediction model according to an embodiment of the present disclosure. Figure 3 ;

[0045] Figure 5 This is a schematic diagram of the structure of a prediction model according to an embodiment of the present disclosure. Figure 4 ;

[0046] Figure 6A and Figure 6B The optimization effect on the 60Hz prediction model before and after introducing the convolutional attention module is shown;

[0047] Figure 7A and Figure 7B The optimization effect on the 120FPS prediction model before and after introducing the convolutional attention module is shown;

[0048] Figure 8 This is a schematic diagram of an adjustment device for a display panel according to an embodiment of the present disclosure;

[0049] Figure 9 This is a block diagram of an electronic device used to implement the adjustment method of the display panel according to the embodiments of the present disclosure. Detailed Implementation

[0050] In the following description, only certain exemplary embodiments are briefly described. As those skilled in the art will recognize, the described embodiments can be modified in various ways without departing from the spirit or scope of this application. Therefore, the drawings and description are considered to be exemplary in nature and not restrictive.

[0051] In the description of this specification, it should be understood that the terms "center," "longitudinal," "transverse," "length," "width," "thickness," "upper," "lower," "front," "rear," "left," "right," "vertical," "horizontal," "top," "bottom," "inner," "outer," "clockwise," "counterclockwise," "axial," "radial," and "circumferential" indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings. They are used only for the convenience of describing this application and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, they should not be construed as limitations on this application.

[0052] Furthermore, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of technical features indicated. Thus, a feature defined as "first" or "second" may explicitly or implicitly include one or more of that feature. In the description of this application, "multiple" means two or more, unless otherwise explicitly specified.

[0053] Figure 1 This is a flowchart illustrating a method for adjusting a display panel according to an embodiment of the present disclosure, including:

[0054] S110, obtain the first gamma parameter of the display panel at the first refresh rate. The first gamma parameter includes the correspondence between grayscale and brightness in multiple color modes that conform to the standard gamma curve.

[0055] S120, the first gamma parameter is input into the prediction model to obtain the second gamma parameter of the display panel at the second refresh frequency. The prediction model includes a downsampling network structure and an upsampling network structure. The second refresh frequency is different from the first refresh frequency.

[0056] S130, adjusts the display panel according to the second gamma parameter.

[0057] The standard gamma curve is the industry-standard gamma value curve, and the currently recognized industry standard gamma value curve is the Gamma 2.2 curve. The first gamma parameter at the first refresh rate can be a manually adjusted parameter conforming to the Gamma 2.2 curve. Multiple color modes can include red mode, blue mode, and green mode. For each color mode, the first gamma parameter includes the brightness value corresponding to each grayscale level. For example, for a specific screen that supports 60Hz and 120Hz display frequencies and contains 16 grayscale levels and 10 brightness values, at a display frequency of 120Hz, for red mode, grayscale 255 corresponds to a brightness value of 5 on the Gamma 2.2 curve; grayscale 111 corresponds to a brightness value of 1 on the Gamma 2.2 curve. For blue mode, grayscale 255 corresponds to a brightness value of 6 on the Gamma 2.2 curve; grayscale 111 corresponds to a brightness value of 4 on the Gamma 2.2 curve. For green mode, grayscale 255 corresponds to a brightness value of 8 on the Gamma 2.2 curve; grayscale 111 corresponds to a brightness value of 1 on the Gamma 2.2 curve. In blue mode, grayscale 255 corresponds to a brightness value of 2 on the Gamma 2.2 curve. Therefore, for this specific screen, its first refresh rate is 120Hz, and the first gamma parameter includes the correspondence between grayscale and brightness for red, blue, and green modes at a frequency of 120Hz.

[0058] The prediction model can be a pre-trained neural network model that uses training data at a first refresh rate of a specific screen through repeated iterations. The trained prediction model can predict the second gamma parameter of the specific screen at a second refresh rate, based on the first gamma parameter that conforms to the Gamma 2.2 curve at the first refresh rate, and output the second gamma parameter for adjusting the specific screen.

[0059] Using the method described in the above embodiments, for screens that support multiple frequency modes, the Gamma curve can be adjusted in only one frequency mode, and the adjusted gamma parameters and the pre-trained prediction model can be used to predict the gamma parameters in other frequency modes. This allows for the adjustment of the Gamma curve of screens with multiple refresh rates to be achieved by only spending the time of adjusting the Gamma curve of a single frequency mode screen, thereby saving adjustment time by a factor of two, improving screen production efficiency, and saving costs.

[0060] In one implementation, Figure 2 This demonstrates a neural network whose downsampling and upsampling network structures are completely corresponding, with the prediction model being... Figure 2 In the case of the neural network shown, step S120 includes:

[0061] The first gamma parameter is subjected to N sets of downsampling operations in an iterative manner through a downsampling network structure to obtain the target downsampling result, where N is a positive integer, and each set of downsampling operations includes convolution and pooling operations;

[0062] The upsampling network structure iteratively performs N sets of upsampling operations on the target downsampling result to obtain the target upsampling result. Each set of upsampling operations includes a deconvolution operation and a feature fusion operation. The feature map corresponding to the deconvolution result of the deconvolution operation is the same size as the feature map corresponding to the convolution result of the corresponding set of downsampling operations. The feature fusion operation is used to fuse the features of the deconvolution result and the convolution result of the corresponding set of downsampling operations.

[0063] The second gamma parameter is obtained based on the target upsampling result.

[0064] For example, the number of convolution operations included in each set of downsampling operations can be configured to be at least one as needed. Figure 2The diagram illustrates a single convolution operation per downsampling operation. The number of downsampling network groups and the number of upsampling network groups can be determined based on actual needs, as long as the number of groups in both is equal. This application is not limited to this. The deconvolution operation in the upsampling operation can use a skip-level network structure to sample the feature map corresponding to the convolution result of the downsampling operation with the same feature map size to extract features. Furthermore, feature fusion operations can be used to fuse features of different depths, outputting a more refined result.

[0065] In one implementation, such as Figure 3 As shown, the upsampling network structure includes a first upsampling network structure and a second upsampling network structure. In this case, step S120 includes:

[0066] The first gamma parameter is subjected to M sets of downsampling operations in an iterative manner through a downsampling network structure to obtain the target downsampling result, where M is an integer greater than 1, and each set of downsampling operations includes convolution and pooling operations.

[0067] The first upsampling network structure iteratively performs M-1 sets of first upsampling operations on the target downsampling result to obtain the first upsampling result. Each set of first upsampling operations includes a deconvolution operation and a feature fusion operation. The feature map corresponding to the deconvolution result obtained by the deconvolution operation is the same size as the feature map corresponding to the convolution result obtained by the corresponding set of downsampling operations. The feature fusion operation is used to fuse the features of the deconvolution result obtained by the deconvolution operation and the convolution result obtained by the corresponding set of downsampling operations.

[0068] The second upsampling network structure performs a second upsampling operation on the first upsampling result to obtain a second upsampling result. The second upsampling operation includes an upsampling operation and a convolution operation.

[0069] The second gamma parameter is obtained based on the second upsampling result.

[0070] For example, the number of groups of first upsampling operations performed by the first upsampling network structure is one less than the number of groups of downsampling operations performed by the downsampling network structure. After performing M-1 groups of first upsampling operations and obtaining the first upsampling result, the subsequent second upsampling operation only performs upsampling and convolution, without performing deconvolution. The significance of setting the network structure in this way is that if the number of deconvolution operations is large, the feature map output by the neural network structure will be large. In this case, continuing to perform deconvolution will reduce the number of channels in the feature map and reduce the features. Therefore, after performing a certain number of deconvolution operations, only upsampling is performed to avoid reducing the output features.

[0071] In one implementation, such as Figure 4As shown, the prediction model also includes a convolutional attention module. In this case, step S120 includes:

[0072] The first downsampling result is obtained by iteratively performing N sets of downsampling operations on the first gamma parameter through a downsampling network structure, where N is a positive integer and each set of downsampling operations includes convolution and pooling operations.

[0073] The first downsampling result is processed by a convolutional attention module to obtain the second downsampling result;

[0074] The upsampling network structure iteratively performs N sets of upsampling operations on the second downsampling result to obtain the target upsampling result. Each set of upsampling operations includes deconvolution and feature fusion operations. The feature map corresponding to the deconvolution result obtained by the deconvolution operation is the same size as the feature map corresponding to the convolution result obtained by the corresponding set of downsampling operations. The feature fusion operation is used to fuse the features of the deconvolution result obtained by the deconvolution operation and the convolution result obtained by the corresponding set of downsampling operations.

[0075] The second gamma parameter is obtained based on the target upsampling result.

[0076] For example, the convolutional attention module combines channel attention and spatial attention mechanisms, which can increase the weights of feature values ​​in deep network layers, thereby improving the predictive model's ability to recognize and extract features. The channel attention mechanism first performs global average pooling and global max pooling on the feature map corresponding to the first downsampling result. Then, based on the sum of the average pooling and max pooling results, a specific function mapping is used to obtain the weight value of each channel of the input feature map, and then the weight value is multiplied by the original input feature map. The spatial attention mechanism calculates the maximum and average values ​​for each element in the feature map, stacks the obtained maximum and average values, adjusts the number of channels using a single convolution with 1 channel, and obtains the weight value of each channel of the input feature map using a specific function mapping. Then, the weight value is multiplied by the original input feature map to output the second downsampling result. Compared to the original first downsampling result, the second downsampling result output after adding the convolutional attention module extracts more refined and accurate features, facilitating subsequent upsampling operations and improving the accuracy of the model's output.

[0077] In one implementation, step S110 includes:

[0078] The first gamma parameter is input into the prediction model in the form of a1×b1×c1, where a1 is the number of brightness levels of the display panel, b1 is the number of gray levels of the display panel, and c1 is the number of color modes of the display panel.

[0079] For example, taking a specific screen supporting 60Hz and 120Hz display frequencies with 16 gray levels and 10 brightness values ​​as an example, assume the first gamma parameter is a pre-adjusted parameter at a 120Hz refresh rate. For this specific screen, since it has three color modes: red, blue, and green, the data form of its input prediction model is a three-dimensional matrix of brightness value × gray level × number of channels. Corresponding to the previous example, the data form of the input prediction model is a 10×16×3 three-dimensional matrix, that is, a feature map of size 10×16 with 3 channels.

[0080] In one implementation, the sampling result obtained by the downsampling operation or the upsampling operation includes a×b×c feature data, where a, b, and c are all positive integers, a represents the height of the feature map obtained by performing the downsampling operation or the upsampling operation, b represents the width of the feature map obtained by performing the downsampling operation or the upsampling operation, and c represents the number of channels of the feature map obtained by performing the downsampling operation or the upsampling operation.

[0081] Figure 5 A prediction model for predicting gamma parameters of a multi-frequency display screen with 16 gray levels and 10 brightness values ​​is shown, such as... Figure 5 As shown, the prediction model includes a downsampling network structure, a convolutional attention module (CBAM), a first upsampling network structure, and a second upsampling network structure.

[0082] The downsampling network structure includes three sets of downsampling operations. Each set of downsampling operations includes one convolution operation and one pooling operation. The adjusted first gamma parameter at 120Hz frequency can be input into the downsampling network structure in the form of a 10×16×3 three-dimensional matrix. After the first convolution operation, the number of channels of the feature map is increased, and the data becomes 10×16×32. Then, the first pooling operation reduces the size of the feature map, and the data becomes 5×8×32. After two more sets of downsampling operations, the first downsampling result output is a feature map with a size of 1×2 and 128 channels.

[0083] Then, the first downsampling result is processed by CBAM to further extract features, but the size and number of channels of the feature map are not changed. The output second downsampling result is still a feature map with a size of 1×2 and 128 channels.

[0084] The first upsampling network structure includes two sets of upsampling operations. Each set of upsampling operations includes one deconvolution operation and one feature fusion operation. The second downsampling result is processed by the two upsampling operations of the first upsampling network structure. The output first upsampling result is a feature map with a size of 5×8 and 32 channels. In this case, the number of channels in the output feature map of the first upsampling result is already relatively small. Therefore, the subsequent second upsampling network structure does not include a deconvolution operation. It only extracts features through the upsampling operation, restores the size of the feature map to the original input size of 10×16, and finally restores the number of channels of the feature map to the original number of 3 channels through a convolution operation, and outputs the predicted second gamma parameter at a frequency of 60Hz.

[0085] It is understood that the descriptions of the number of operations such as convolution, deconvolution, and pooling, as well as the descriptions of the data format in the various embodiments of this application, are only illustrative descriptions for specific screens. In practical applications, corresponding prediction model structures can be designed and trained according to different screen physical properties, and this application is not limited thereto.

[0086] Furthermore, the prediction model in this embodiment can output the predicted gamma parameters of that specific screen at another refresh rate, conforming to the Gamma 2.2 curve, by inputting the gamma parameters of a specific screen at one adjusted refresh rate. However, for screens supporting three or more refresh rates, to ensure the accuracy of the model's prediction, a corresponding additional prediction model needs to be trained for each additional refresh rate. For example, for 60Hz and 120PFS frequency modes, two different prediction models are trained using the first gamma parameter at 120Hz as samples.

[0087] In one implementation, the model can be evaluated by calculating the mean absolute error (MAE) between the model's predicted values ​​and the true values, in order to verify the optimization effect of adding the convolutional attention module on the model.

[0088] Figure 6A and Figure 6B The absolute errors of the second gamma parameter predicted by the 60Hz prediction model before and after the introduction of the convolutional attention module are shown.

[0089] like Figure 6A As shown, without the introduction of a convolutional attention module, among the 100 test panels, the maximum absolute error between the predicted and true values ​​was 2.09, the minimum was 0.99, and the average absolute error was 1.36. Figure 6BIn the test, after introducing the convolutional attention module, the maximum absolute error between the predicted value and the true value was 1.98, the minimum was 1.30, and the average absolute error was 1.30. Thus, compared with before introducing the convolutional attention module, the average absolute error between the model prediction result and the true value was reduced by 0.06, which means that the convolutional attention module is effective in optimizing the 60Hz prediction model.

[0090] Figure 7A and Figure 7B The absolute errors of the second gamma parameter predicted by the 120FPS prediction model before and after the introduction of the convolutional attention module are shown respectively, compared with the true value of the gamma parameter conforming to the Gamma2.2 curve.

[0091] like Figure 7A As shown, without the introduction of a convolutional attention module, among the 100 test panels, the maximum absolute error between the predicted and true values ​​was 2.30, the minimum was 0.96, and the average absolute error was 1.31. Figure 7B In the test, after introducing the convolutional attention module, the maximum absolute error between the predicted value and the true value was 1.87, the minimum was 0.76, and the average absolute error was 1.19 in 100 panels. Thus, compared with before introducing the convolutional attention module, the average absolute error between the model prediction result and the true value was reduced by 0.12, which means that the convolutional attention module is effective in optimizing the 120FPS prediction model.

[0092] The specific settings and implementation methods of the embodiments of this application have been described above from different perspectives. Using the methods provided in the above embodiments, various prediction models with different network structures are provided, all used to predict the gamma parameters of a specific screen in one frequency mode based on the gamma parameters in another frequency mode. This allows for the adjustment of the gamma curve of screens with multiple refresh rates to be achieved using only the gamma curve adjustment time for a single frequency mode screen, significantly saving adjustment time and improving screen production efficiency.

[0093] As an implementation of the above methods, such as Figure 8 As shown in the embodiments of this disclosure, an adjustment device for a display panel is also provided, comprising:

[0094] The acquisition module 810 is used to acquire the first gamma parameter of the display panel at the first refresh rate. The first gamma parameter includes the correspondence between grayscale and brightness in multiple color modes that conform to the standard gamma curve.

[0095] The input module 820 is used to input the first gamma parameter into the prediction model to obtain the second gamma parameter of the display panel at the second refresh frequency. The prediction model includes a downsampling network structure and an upsampling network structure, and the second refresh frequency is different from the first refresh frequency.

[0096] Adjustment module 830 is used to adjust the display panel according to the second gamma parameter.

[0097] For example, the input module 820 includes:

[0098] The downsampling unit is used to perform N sets of downsampling operations on the first gamma parameter in an iterative manner through the downsampling network structure to obtain the target downsampling result, where N is a positive integer, and each set of downsampling operations includes convolution and pooling operations;

[0099] The upsampling unit is used to iteratively perform N sets of upsampling operations on the target downsampling result through the upsampling network structure to obtain the target upsampling result. Each set of upsampling operations includes a deconvolution operation and a feature fusion operation. The feature map corresponding to the deconvolution result obtained by the deconvolution operation is the same size as the feature map corresponding to the convolution result obtained by the corresponding set of downsampling operations. The feature fusion operation is used to fuse the features of the deconvolution result obtained by the deconvolution operation and the convolution result obtained by the corresponding set of downsampling operations.

[0100] The parameter acquisition unit is used to obtain the second gamma parameter based on the target upsampling result.

[0101] Optionally, the upsampling network structure includes a first upsampling network structure and a second upsampling network structure, and the input module 820 is used for:

[0102] The first gamma parameter is subjected to M sets of downsampling operations in an iterative manner through a downsampling network structure to obtain the target downsampling result, where M is an integer greater than 1, and each set of downsampling operations includes convolution and pooling operations.

[0103] The first upsampling network structure iteratively performs M-1 sets of first upsampling operations on the target downsampling result to obtain the first upsampling result. Each set of first upsampling operations includes a deconvolution operation and a feature fusion operation. The feature map corresponding to the deconvolution result obtained by the deconvolution operation is the same size as the feature map corresponding to the convolution result obtained by the corresponding set of downsampling operations. The feature fusion operation is used to fuse the features of the deconvolution result obtained by the deconvolution operation and the convolution result obtained by the corresponding set of downsampling operations.

[0104] The second upsampling network structure performs a second upsampling operation on the first upsampling result to obtain a second upsampling result. The second upsampling operation includes an upsampling operation and a convolution operation.

[0105] The second gamma parameter is obtained based on the second upsampling result.

[0106] Optionally, the prediction model also includes a convolutional attention module, with input module 820 used for:

[0107] The first downsampling result is obtained by iteratively performing N sets of downsampling operations on the first gamma parameter through a downsampling network structure, where N is a positive integer and each set of downsampling operations includes convolution and pooling operations.

[0108] The first downsampling result is processed by a convolutional attention module to obtain the second downsampling result;

[0109] The upsampling network structure iteratively performs N sets of upsampling operations on the second downsampling result to obtain the target upsampling result. Each set of upsampling operations includes deconvolution and feature fusion operations. The feature map corresponding to the deconvolution result obtained by the deconvolution operation is the same size as the feature map corresponding to the convolution result obtained by the corresponding set of downsampling operations. The feature fusion operation is used to fuse the features of the deconvolution result obtained by the deconvolution operation and the convolution result obtained by the corresponding set of downsampling operations.

[0110] The second gamma parameter is obtained based on the target upsampling result.

[0111] Optionally, the acquisition module 810 is used for:

[0112] The first gamma parameter is input into the prediction model in the form of a1×b1×c1, where a1 is the number of brightness levels of the display panel, b1 is the number of gray levels of the display panel, and c1 is the number of color modes of the display panel.

[0113] Optionally, the sampling result obtained by the downsampling or upsampling operation in the above-mentioned device includes a×b×c feature data, where a, b, and c are all positive integers, a represents the height of the feature map obtained by performing the downsampling or upsampling operation, b represents the width of the feature map obtained by performing the downsampling or upsampling operation, and c represents the number of channels of the feature map obtained by performing the downsampling or upsampling operation.

[0114] The functions of each unit, module, or sub-module in the various devices of this disclosure embodiment can be found in the corresponding descriptions in the above method embodiments, and they have corresponding beneficial effects, which will not be repeated here.

[0115] The acquisition, storage, and application of user personal information involved in the technical solution disclosed herein comply with the provisions of relevant laws and regulations and do not violate public order and good morals.

[0116] Figure 9 A structural block diagram of an electronic device according to an embodiment of this application is shown. Figure 9As shown, the electronic device includes a memory 910 and a processor 920. The memory 910 stores instructions that can be executed on the processor 920. When the processor 920 executes the instructions, it implements the methods described in the above embodiments. The number of memories 910 and processors 920 can be one or more. This electronic device is intended to represent various forms of digital computers, such as laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device can also represent various forms of mobile devices, such as personal digital processors, cellular phones, smartphones, wearable devices, and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely examples and are not intended to limit the implementation of the present application described and / or claimed herein.

[0117] The electronic device may also include a communication interface 930 for communicating with external devices and exchanging data. The devices are interconnected using different buses and can be mounted on a common motherboard or otherwise as needed. The processor 920 can process instructions executed within the electronic device, including instructions stored in or on memory to display graphical information of a GUI on an external input / output device (such as a display device coupled to the interface). In other embodiments, multiple processors and / or multiple buses can be used with multiple memories and multiple memory modules, if desired. Similarly, multiple electronic devices can be connected, each providing some of the necessary operations (e.g., as a server array, a group of blade servers, or a multiprocessor system). The bus can be divided into address buses, data buses, control buses, etc. For ease of illustration, Figure 9 The bus is represented by a single thick line, but this does not mean that there is only one bus or one type of bus.

[0118] Optionally, in a specific implementation, if the memory 910, processor 920, and communication interface 930 are integrated on a single chip, the memory 910, processor 920, and communication interface 930 can communicate with each other through an internal interface.

[0119] It should be understood that the aforementioned processor can be a Central Processing Unit (CPU), or other general-purpose processors, Digital Signal Processors (DSPs), Application Specific Integrated Circuits (ASICs), Field-Programmable Gate Arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. General-purpose processors can be microprocessors or any conventional processor. It is worth noting that the processor can be a processor supporting Advanced Reduced Instruction Set Machines (ARM) architecture.

[0120] This application provides a non-transient computer-readable storage medium (such as the memory 910 described above) that stores computer instructions that, when executed by a processor, implement the method provided in this application.

[0121] Optionally, the memory 910 may include a program storage area and a data storage area, wherein the program storage area may store the operating system and applications required for at least one function; the data storage area may store data created based on the use of the electronic device for recognizing lane edges, etc. Furthermore, the memory 910 may include high-speed random access memory and may also include non-transitory memory, such as at least one disk storage device, flash memory device, or other non-transitory solid-state storage device. In some embodiments, the memory 910 may optionally include memory remotely generated relative to the processor 920, which can be connected to the electronic device for recognizing lane edges via a network. Examples of such networks include, but are not limited to, the Internet, corporate intranets, local area networks, mobile communication networks, and combinations thereof.

[0122] Computer-readable media include both permanent and non-permanent, removable and non-removable media that can store information by any method or technology. Information can be computer-readable instructions, data structures, modules of programs, or other data. Examples of computer storage media include, but are not limited to, phase-change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other physical classes of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technologies, CD-ROM, digital versatile optical disc (DVD) or other optical storage, magnetic cassettes, magnetic disk storage or other magnetic storage media, or any other non-transferable media that can be used to store information accessible by a computing device. As defined herein, computer-readable media does not include non-transitory computer-readable media, such as modulated data signals and carrier waves.

[0123] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any person skilled in the art can easily conceive of various variations or substitutions within the technical scope disclosed in this application, and these should all be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

Claims

1. A method for adjusting a display panel, characterized in that, include: Obtain the first gamma parameter of the display panel at the first refresh rate. The first gamma parameter includes the correspondence between grayscale and brightness in multiple color modes that conform to the standard gamma curve. The first gamma parameter is input into the prediction model to obtain the second gamma parameter of the display panel at the second refresh frequency. The prediction model includes a downsampling network structure and an upsampling network structure. The second refresh frequency is different from the first refresh frequency. The step of inputting the first gamma parameter into the prediction model includes: The first gamma parameter is input into the prediction model in the form of a1×b1×c1, where a1 is the number of brightness levels of the display panel, b1 is the number of gray levels of the display panel, and c1 is the number of color modes of the display panel; the downsampling network structure includes the same number of downsampling operations as the upsampling network structure. The display panel is adjusted according to the second gamma parameter.

2. The method according to claim 1, characterized in that, The step of inputting the first gamma parameter into the prediction model to obtain the second gamma parameter of the display panel at the second refresh rate includes: The first gamma parameter is subjected to N sets of downsampling operations in an iterative manner through the downsampling network structure to obtain the target downsampling result, where N is a positive integer, and each set of downsampling operations includes convolution and pooling operations; The upsampling network structure iteratively performs N sets of upsampling operations on the target downsampling result to obtain the target upsampling result. Each set of upsampling operations includes a deconvolution operation and a feature fusion operation. The feature map corresponding to the deconvolution result obtained by the deconvolution operation has the same size as the feature map corresponding to the convolution result obtained by the corresponding set of downsampling operations. The feature fusion operation is used to fuse the features of the deconvolution result obtained by the deconvolution operation and the convolution result obtained by the corresponding set of downsampling operations. The second gamma parameter is obtained based on the target upsampling result.

3. The method according to claim 1, characterized in that, The upsampling network structure includes a first upsampling network structure and a second upsampling network structure. The step of inputting the first gamma parameter into the prediction model to obtain the second gamma parameter of the display panel at the second refresh rate includes: The first gamma parameter is subjected to M sets of downsampling operations in an iterative manner through the downsampling network structure to obtain the target downsampling result, where M is an integer greater than 1, and each set of downsampling operations includes convolution and pooling operations. The first upsampling network structure iteratively performs M-1 sets of first upsampling operations on the target downsampling result to obtain the first upsampling result. Each set of first upsampling operations includes a deconvolution operation and a feature fusion operation. The feature map corresponding to the deconvolution result obtained by the deconvolution operation has the same size as the feature map corresponding to the convolution result obtained by the corresponding set of downsampling operations. The feature fusion operation is used to fuse the features of the deconvolution result obtained by the deconvolution operation and the convolution result obtained by the corresponding set of downsampling operations. The second upsampling network structure is used to perform a second upsampling operation on the first upsampling result to obtain a second upsampling result. The second upsampling operation includes an upsampling operation and a convolution operation. The second gamma parameter is obtained based on the second upsampling result.

4. The method according to claim 1, characterized in that, The prediction model further includes a convolutional attention module. The step of inputting the first gamma parameter into the prediction model to obtain the second gamma parameter of the display panel at the second refresh rate includes: The first downsampling result is obtained by iteratively performing N sets of downsampling operations on the first gamma parameter through the downsampling network structure, where N is a positive integer and each set of downsampling operations includes convolution and pooling operations. The first downsampling result is processed by the convolutional attention module to obtain the second downsampling result; The upsampling network structure iteratively performs N sets of upsampling operations on the second downsampling result to obtain the target upsampling result. Each set of upsampling operations includes a deconvolution operation and a feature fusion operation. The feature map corresponding to the deconvolution result obtained by the deconvolution operation is the same size as the feature map corresponding to the convolution result obtained by the corresponding set of downsampling operations. The feature fusion operation is used to fuse the features of the deconvolution result obtained by the deconvolution operation and the convolution result obtained by the corresponding set of downsampling operations. The second gamma parameter is obtained based on the target upsampling result.

5. The method according to any one of claims 2-4, characterized in that, The sampling result obtained by the downsampling operation or the upsampling operation includes a×b×c feature data, where a, b, and c are all positive integers, a represents the height of the feature map obtained by performing the downsampling operation or the upsampling operation, b represents the width of the feature map obtained by performing the downsampling operation or the upsampling operation, and c represents the number of channels of the feature map obtained by performing the downsampling operation or the upsampling operation.

6. An adjustment device for a display panel, characterized in that, include: The acquisition module is used to acquire the first gamma parameter of the display panel at the first refresh rate. The first gamma parameter includes the correspondence between grayscale and brightness in multiple color modes that conform to the standard gamma curve. An input module is used to input the first gamma parameter into a prediction model to obtain the second gamma parameter of the display panel at a second refresh frequency. The prediction model includes a downsampling network structure and an upsampling network structure, and the second refresh frequency is different from the first refresh frequency. Inputting the first gamma parameter into the prediction model includes: inputting the first gamma parameter into the prediction model in the form of a1×b1×c1, where a1 is the number of brightness levels of the display panel, b1 is the number of grayscale levels of the display panel, and c1 is the number of color modes of the display panel; the downsampling network structure includes an equal number of downsampling operations as the upsampling network structure. An adjustment module is used to adjust the display panel according to the second gamma parameter.

7. The apparatus according to claim 6, characterized in that, The input module includes: The downsampling unit is used to perform N sets of downsampling operations on the first gamma parameter in an iterative manner through the downsampling network structure to obtain the target downsampling result, where N is a positive integer, and each set of downsampling operations includes convolution and pooling operations; An upsampling unit is used to iteratively perform N sets of upsampling operations on the target downsampling result through the upsampling network structure to obtain the target upsampling result. Each set of upsampling operations includes a deconvolution operation and a feature fusion operation. The feature map corresponding to the deconvolution result obtained by the deconvolution operation has the same size as the feature map corresponding to the convolution result obtained by the corresponding set of downsampling operations. The feature fusion operation is used to fuse the features of the deconvolution result obtained by the deconvolution operation and the convolution result obtained by the corresponding set of downsampling operations. The parameter acquisition unit is used to obtain the second gamma parameter based on the target upsampling result.

8. An electronic device, characterized in that, include: At least one processor; as well as A memory communicatively connected to the at least one processor; wherein, The memory stores instructions that can be executed by the at least one processor to enable the at least one processor to perform the method of any one of claims 1-5.

9. A non-transient computer-readable storage medium having a computer program stored thereon, characterized in that, When the program is executed by the processor, it implements the method of any one of claims 1 to 5.

Images