Compensation parameter adjusting method, adjusting device, display device and program product

Abstract

The invention provides a compensation parameter adjusting method and device, a display device and a program product, and is applied to the technical field of display. The method comprises the following steps: acquiring an interface image displayed by a first display screen; performing data analysis on the interface image of the first display screen to obtain a quantitative deviation value of the interface image of the first display screen relative to a corresponding standard interface; inputting the quantization deviation value into a pre-trained AI model, and obtaining a Demura parameter for performing display compensation on the first display screen according to the output of the AI model; the AI model is obtained by training a plurality of acquisition quantization deviation values and a plurality of compensation parameters. By adopting the method, the adjustment of the Demura parameter can be automatically completed, manual implementation is not needed, and independent adjustment can be realized for each display screen in a batch production scale.

Description

Technical Field

[0001] The present invention relates to the field of display technology, and in particular to a compensation parameter adjustment method, an adjustment device, a display device and a program product. Background Art

[0002] During the manufacturing process of Organic Light-Emitting Diode (OLED) display screens, there are cases where the relevant parameters of the light-emitting units deviate from the design values, resulting in a deviation between the luminous brightness of the display screen and the ideal value, which in turn causes the display effect to deteriorate. This phenomenon is usually called Mura (uneven display phenomenon).

[0003] The presence of Mura will reduce the display quality of the display screen and thus lead to a decrease in product yield. Therefore, the Demura (Mura elimination) method needs to be adopted in the production process of the display screen to externally compensate for the Mura defects that appear in the production of the display screen, reduce or eliminate the Mura defects, and improve the display effect and yield of the product.

[0004] The Demura method of the prior art usually uses manual adjustment to determine the Demura parameters used for Mura defect compensation, which is not only time-consuming and labor-intensive, but also difficult to achieve the desired effect, and is not suitable for independent adjustment in mass production. Summary of the invention

[0005] The technical solution of the present invention aims to provide a compensation parameter adjustment method, an adjustment device, a display device and a program product, which are used to solve the problem that the Demura parameters for Mura compensation are manually adjusted in the prior art, which is time-consuming and labor-intensive to achieve an ideal display effect, and is not suitable for independent adjustment in mass production.

[0006] One embodiment of the present invention provides a compensation parameter adjustment method, which includes:

[0007] Acquiring an interface image displayed on a first display screen; wherein a mura defect is displayed on the interface image;

[0008] Performing data analysis on the interface image of the first display screen to obtain a quantitative deviation value of the interface image of the first display screen compared with a corresponding standard interface;

[0009] The quantization deviation value is input into a pre-trained AI model, and the Demura parameters for display compensation of the first display screen are obtained according to the output of the AI ​​model; wherein the AI ​​model is obtained by training multiple acquired quantization deviation values ​​and multiple compensation parameters, the acquired quantization deviation value is the quantization deviation value between the acquired interface images of the multiple second display screens and the corresponding standard interface, and the compensation parameter is the parameter required to compensate for adjusting the acquired interface image to the corresponding standard interface.

[0010] Optionally, the compensation parameter adjustment method further comprises:

[0011] According to the Demura parameters and the pre-obtained mura data of the first display screen, the display driving parameters of the first display screen are adjusted to obtain adjusted display parameters;

[0012] According to the adjusted display parameters, the interface image presented by the first display screen is adjusted.

[0013] Optionally, the compensation parameter adjustment method further comprises:

[0014] Returning to the step of reacquiring the interface image displayed on the first display screen until the obtained quantization deviation value is less than or equal to a preset deviation value;

[0015] Determine that when the quantization deviation value is less than or equal to the preset deviation value, the corresponding quantization deviation value is input into the Demura parameter obtained by the AI ​​model, which is the final Demura parameter of the first display screen.

[0016] Optionally, the compensation parameter adjustment method further comprises:

[0017] Performing filtering and / or edge processing on the interface image presented on the first display screen to obtain a processed interface image;

[0018] Obtaining mura data of the first display screen by performing data analysis on the processed interface image;

[0019] The mura data is stored in a memory.

[0020] Optionally, the compensation parameter adjustment method further comprises:

[0021] Performing data analysis on the obtained multiple interface images of the multiple second display screens to obtain a quantitative deviation value of each interface image of the second display screen compared with a corresponding standard interface;

[0022] Calculating compensation data corresponding to the interface image of the second display screen according to each interface image of the plurality of second display screens collected and the corresponding standard interface;

[0023] Using a principal component analysis (PCA) learning algorithm to determine basic compensation data from a plurality of compensation data;

[0024] The quantization deviation value corresponding to the interface images of the multiple second display screens is set as the acquisition quantization deviation value, and the acquisition quantization deviation value is used as the input of the model training, and the basic compensation data is used as the output of the model training, and the gradient descent algorithm is used for training to obtain the AI ​​model.

[0025] Optionally, the compensation parameter adjustment method, wherein data analysis is performed on the interface image to obtain a quantitative deviation value of the interface image compared to a corresponding standard interface, comprises:

[0026] Performing data analysis on the interface image to obtain an effect quantification value corresponding to the interface image; the effect quantification value is a characterization value used to indicate the severity of the mura defect;

[0027] The effect quantization value corresponding to the interface image is compared with the effect quantization value corresponding to the standard interface to obtain the quantization deviation value.

[0028] Optionally, the compensation parameter adjustment method, wherein data analysis is performed on the interface image to obtain a quantitative effect value corresponding to the interface image, comprises:

[0029] After dividing the interface image into a plurality of sub-regions, the number of mura defects in each sub-region is sequentially obtained using an image processing algorithm;

[0030] Determining a defect characterization value corresponding to the interface image according to the total number of mura defects in the multiple sub-regions and the number of the multiple sub-regions;

[0031] According to the defect characterization value, the corresponding effect quantification value is determined.

[0032] Optionally, the compensation parameter adjustment method further comprises:

[0033] Performing simulation processing on the interface image to obtain a standard interface corresponding to the interface image;

[0034] Set the effect quantization value corresponding to the standard interface to the preset value.

[0035] One embodiment of the present invention further provides a compensation parameter adjustment device, which includes:

[0036] An image acquisition module, used for acquiring an interface image displayed on the first display screen; the interface image shows a mura defect;

[0037] An analysis module, configured to perform data analysis on the interface image of the first display screen to obtain a quantitative deviation value of the interface image of the first display screen compared with a corresponding standard interface;

[0038] A processing module is used to input the quantization deviation value into a pre-trained AI model, and obtain Demura parameters for display compensation of the display screen according to the output of the AI ​​model; wherein the AI ​​model is obtained by training multiple acquired quantization deviation values ​​and multiple compensation parameters, the acquired quantization deviation value is the quantization deviation value between one of the interface images of the acquired multiple second display screens and the corresponding standard interface, and the compensation parameter is the parameter required to compensate for adjusting the acquired interface image to the corresponding standard interface.

[0039] One embodiment of the present invention further provides a display device, which includes a processor, a memory, and a program stored in the memory and executable on the processor, wherein the program, when executed by the processor, implements the compensation parameter adjustment method as described in any one of the above items.

[0040] One embodiment of the present invention further provides a computer program product, which includes computer instructions, and when the computer instructions are executed by a processor, the steps in the compensation parameter adjustment method as described above are implemented.

[0041] At least one of the above technical solutions in the specific embodiment of the present invention has the following beneficial effects:

[0042] By adopting the compensation parameter adjustment method described in the embodiment of the present invention, an AI model is obtained by training and deducing the quantized deviation value of the interface image of the display screen compared with the standard interface and the compensation parameter of the interface image compared with the standard interface, and the quantized deviation value of the collected interface image of the first display screen compared with the corresponding standard interface is analyzed to obtain the Demura parameters for display compensation of the display screen. In this way, by adopting this method, the adjustment of the Demura parameters can be completed automatically without manual implementation, and for each display screen in the mass production scale, individual adjustment can be achieved, thereby effectively ensuring the best display adjustment effect of each display screen. BRIEF DESCRIPTION OF THE DRAWINGS

[0043] Figure 1 A schematic diagram of a system architecture using the compensation parameter adjustment method according to an embodiment of the present invention;

[0044] Figure 2 A schematic diagram of a flow chart of a compensation parameter adjustment method according to an embodiment of the present invention;

[0045] Figure 3 A schematic diagram of a flow chart of one implementation of the method described in an embodiment of the present invention;

[0046] Figure 4 It is a schematic diagram of the structure of the compensation parameter adjustment device described in an embodiment of the present invention. DETAILED DESCRIPTION

[0047] In order to make the technical problems, technical solutions and advantages to be solved by the present invention more clear, a detailed description will be given below with reference to the accompanying drawings and specific embodiments.

[0048] During the production process of OLED display screens, due to reasons such as materials and processes, some products may have uneven screen brightness, which is called Mura. Such uneven brightness spots on the display screen will cause visual discomfort. At present, the external compensation method to solve this problem includes: sensing the electrical or optical characteristics of pixels through external driving circuits or devices to detect mura, and compensating and eliminating mura after detecting mura, i.e., demura, and determining demura parameters. The determined demura parameters are written into the flash memory Flash ROM of the display screen to adjust the display parameters of the real-time output display screen so that the output display screen eliminates mura.

[0049] Demura parameters used for Mura defect compensation are usually determined by the Demura method of the prior art and are usually adjusted manually, which is not only time-consuming and labor-intensive, but also difficult to achieve the desired effect. Moreover, due to the scale of mass production, it is unrealistic to adjust each display screen individually. Usually, the same set of parameters are used for the same batch of products. This processing method cannot effectively guarantee the best display adjustment effect for each display screen, and reduces the compensation ability of the Demura algorithm. To solve the above technical problems, an embodiment of the present invention provides a compensation parameter adjustment method, which uses an AI model obtained by training and deducing the quantized deviation value of the interface image of the display screen compared with the standard interface and the compensation parameter of the interface image compared with the standard interface, analyzes the quantized deviation value of the interface image of the first display screen collected compared with the corresponding standard interface, and obtains the Demura parameters for display compensation of the display screen. In this way, the adjustment of the Demura parameters can be completed automatically without manual implementation, and each display screen in the mass production scale can be adjusted individually, thereby effectively ensuring the best display adjustment effect of each display screen.

[0050] like Figure 1The system architecture diagram of the compensation parameter adjustment method according to the embodiment of the present invention is shown in FIG. 1 , wherein the system adopting the compensation parameter adjustment method comprises: a display screen 1, a camera 2 and a processor 3. Optionally, the processor 3 can be arranged on the display screen 1 to form one of the processors of the display device.

[0051] In one embodiment, optionally, the processor 3 may also be composed of multiple different module parts, each of which may be formed into a chip, such as an AI chip provided with a pre-trained AI model, a driver chip for outputting a driving display signal, and a chip for performing other data scheduling processing.

[0052] According to the above system architecture, an embodiment of the present invention provides a compensation parameter adjustment method, wherein one implementation mode, optionally, is applied to the above processor 3, and can use a pre-trained AI model to obtain Demura parameters for display compensation of the display screen, such as Figure 2 As shown, the compensation parameter adjustment method described in the embodiment of the present invention includes:

[0053] S201, acquiring an interface image displayed on a first display screen; the interface image displays a mura defect;

[0054] S202, performing data analysis on the interface image of the first display screen to obtain a quantitative deviation value of the interface image of the first display screen compared with a corresponding standard interface;

[0055] S203, inputting the quantization deviation value into a pre-trained AI model, and obtaining the Demura parameters for display compensation of the first display screen according to the output of the AI ​​model; wherein the AI ​​model is obtained by training multiple acquired quantization deviation values ​​and multiple compensation parameters, the acquired quantization deviation value is the quantization deviation value between the acquired interface images of the multiple second display screens and the corresponding standard interfaces, and the compensation parameter is the parameter required to compensate for adjusting the acquired interface image to the corresponding standard interface.

[0056] In an embodiment of the present invention, the first display screen mentioned may be any screen in the display screen preparation process. By adopting the compensation parameter adjustment method, after obtaining an interface image showing a mura defect on the first display screen, a quantized deviation value of the interface image compared with the corresponding standard interface is obtained by performing data analysis on the interface image. The quantized deviation value is used as the input of a pre-trained AI model. The AI ​​model can be used to obtain Demura parameters for display compensation of the first display screen. In this way, the adjustment of the Demura parameters can be completed automatically without manual implementation, and each display screen in the mass production scale can be adjusted individually, thereby effectively ensuring the best display adjustment effect of each display screen.

[0057] It should be noted that, in the embodiment of the present invention, the standard interface of the interface image mentioned is an interface that uses the same display parameters as the corresponding interface image and the image presented does not include mura defects, or in other words, the standard interface is the corresponding interface of the corresponding interface image that eliminates mura defects. Among them, the compensation parameter adjustment method described in the embodiment of the present invention, before implementation, also includes the step of performing parameter training to obtain an AI model, including:

[0058] Performing data analysis on the obtained multiple interface images of the multiple second display screens to obtain a quantitative deviation value of each interface image of the second display screen compared with a corresponding standard interface;

[0059] Calculating compensation data corresponding to the interface image of the second display screen according to each interface image of the plurality of second display screens collected and the corresponding standard interface;

[0060] Using a PCA learning algorithm to determine basic compensation data from the plurality of compensation data;

[0061] The quantization deviation value corresponding to the interface images of the multiple second display screens is set as the acquisition quantization deviation value, and the acquisition quantization deviation value is used as the input of the model training, and the basic compensation data is used as the output of the model training, and the gradient descent algorithm is used for training to obtain the AI ​​model.

[0062] Wherein, the interface images of the plurality of second display screens are collected images for data training. Optionally, the method further includes:

[0063] When a plurality of first display screens are lit up to display images, the images displayed on the second display screens are captured by a camera to obtain interface images of the plurality of second display screens.

[0064] Optionally, the camera can be a CCD camera or an industrial camera with high resolution and high precision. With this implementation, after the display screen is lit, the interface image of the second display screen can be obtained by taking pictures with a CCD camera or an industrial camera. By analyzing the brightness of each pixel on the interface image, a brightness signal corresponding to the interface image can be obtained, and the mura defect corresponding to the interface image can be obtained based on the brightness signal.

[0065] In one embodiment of the present invention, when the second display screen displays an image with a target display parameter, the displayed image is captured by a CCD camera or an industrial camera to obtain an interface image of the second display screen at the target display parameter, wherein the target display parameter is an image parameter that can clearly present mura defects. Optionally, the image presented according to the target display parameter is a grayscale image, but it is not limited to being a grayscale image, and can also be an RGB image.

[0066] Optionally, the target display parameters include but are not limited to only one or more of the following display parameters:

[0067] First display parameters: brightness is 500 nits, grayscale is 32;

[0068] Second display parameters: brightness is 159 nits, grayscale is 128;

[0069] The third display parameter: brightness is 90 nits, grayscale is 255;

[0070] Fourth display parameter: brightness is 2 nits, grayscale is 32.

[0071] By adopting this implementation mode, when multiple second display screens are respectively presented images according to the above-mentioned display parameters of each type, interface images when each type of image is displayed are obtained through the camera, and a large number of interface images with mura defects can be collected for model training to obtain an AI model capable of performing Demura parameter analysis.

[0072] In the embodiment of the present invention, optionally, performing data analysis on the interface image to obtain a quantitative deviation value of the interface image compared with a corresponding standard interface includes:

[0073] Performing data analysis on the interface image to obtain an effect quantification value corresponding to the interface image; the effect quantification value is a characterization value used to indicate the severity of the mura defect;

[0074] The effect quantization value corresponding to the interface image is compared with the effect quantization value corresponding to the standard interface to obtain the quantization deviation value.

[0075] In one implementation manner, optionally, performing data analysis on the interface image to obtain a quantitative effect value corresponding to the interface image includes:

[0076] After dividing the interface image into a plurality of sub-regions, the number of mura defects in each sub-region is sequentially obtained using an image processing algorithm;

[0077] An effect quantification value corresponding to the interface image is determined according to the total number of mura defects in the multiple sub-regions and the number of the multiple sub-regions.

[0078] Specifically, when the display screen presents an image with one of the display parameters (target display parameters, such as grayscale X and brightness Y), the camera can capture the interface image at the target display parameter. In order to accurately obtain the effect quantization value corresponding to the interface image, the interface image can be divided into multiple sub-areas. Optionally, the interface image can be evenly divided into multiple sub-areas, such as m sub-areas. After that, the image processing algorithm can be used to traverse each sub-area in turn to obtain the number of mura defects in each sub-area, and the total number of mura defects in m sub-areas can be obtained by counting, such as i. According to the total number of mura defects in the m sub-areas and the number of multiple sub-areas m, the defect characterization value N corresponding to the interface image can be determined: N = i / m, and the corresponding effect quantization value n can be determined according to the defect characterization value. Specifically, the effect quantization value n is determined as follows: n = 1-N, that is, n = 1-i / m.

[0079] By adopting the above principle, for the interface images presented by the plurality of second display screens with different display parameters, the effect quantization value corresponding to each interface image can be obtained respectively.

[0080] Optionally, the image processing algorithm for obtaining the number of mura defects in each sub-region may include one or more of convolutional neural networks, genetic algorithms, and recurrent neural networks. By using one or more of these algorithms to perform data analysis on each sub-region of the interface image, the situation of mura defects in each sub-region can be obtained.

[0081] In the embodiment of the present invention, the effect quantization value corresponding to the interface image is set to n, optionally, 0≤n≤1, wherein the larger n is, the lighter the severity of the mura defect is, and the better the compensation effect is.

[0082] Optionally, the method further comprises:

[0083] Performing simulation processing on the obtained interface image to obtain a standard interface corresponding to the interface image;

[0084] Set the effect quantization value corresponding to the standard interface to the preset value.

[0085] The standard interface is the interface image corresponding to the interface image for eliminating mura defects. Optionally, the obtained interface image can be used for simulation to obtain an ideal compensation image (standard interface), and the effect quantization value of the determined ideal compensation image is set as the standard value, that is, the preset value. Optionally, when the effect quantization value n is less than or equal to 1 and greater than or equal to 0, the standard value can be 1.

[0086] It should be noted that the value range of the effect quantification value and the value of the standard value are not limited to the above examples, and can be specifically set according to the calculation parameter accuracy and range.

[0087] According to the above process, by obtaining the effect quantization values ​​n corresponding to the multiple interface images of the multiple second display screens, the effect quantization value n corresponding to each interface image is compared with the effect quantization value corresponding to the standard interface (such as 1), and the deviation value between the two is obtained, that is, the quantitative deviation value Δn is obtained as: Δn=1-n.

[0088] After performing data analysis on the obtained multiple interface images of the multiple second display screens and obtaining the quantitative deviation value of each interface image compared to the corresponding standard interface, further, according to each interface image of the multiple second display screens collected and the corresponding standard interface, the compensation data corresponding to the interface image of the second display screen is calculated. Optionally, the Demura algorithm can be used to calculate the compensation data of each interface image relative to the standard interface. It should be noted that the compensation data is theoretical compensation data.

[0089] Furthermore, a principal component analysis (PCA) learning algorithm is used to determine basic compensation data among the plurality of compensation data.

[0090] Among them, the PCA learning algorithm is a dimensionality reduction algorithm, which is usually used for data dimension compression and feature extraction. It maps the original high-dimensional data to a low-dimensional space through linear transformation. The reduced-dimensional data is easier to visualize and analyze, and can achieve the effect of reducing computational complexity, reducing noise, and preventing overfitting.

[0091] By adopting this implementation mode, the PCA learning algorithm is used to perform dimensionality compression and feature extraction on the theoretical compensation data to achieve dimensionality reduction of the compensation data and obtain the compensation data after dimensionality reduction; thereafter, a distribution relationship diagram can be simulated based on the multiple compensation data after dimensionality reduction, and based on the distribution range of the multiple compensation data in the distribution relationship diagram, multiple compensation data with a concentrated distribution range in the multiple compensation data after dimensionality reduction are selected as basic compensation data for subsequent AI model training.

[0092] After using the PCA learning algorithm to determine the basic compensation data among the multiple compensation data, further, the quantization deviation value corresponding to the interface images of the multiple second display screens can be set as the acquisition quantization deviation value, and the acquisition quantization deviation value can be used as the input of the model training, and the basic compensation data can be used as the output of the model training, and the gradient descent algorithm can be used for training to obtain the AI ​​model.

[0093] The gradient descent algorithm is a method for obtaining the value of a minimized function, and the steps when applied to the training model include:

[0094] 1. Initialization parameters: randomly select an initial point;

[0095] 2. Calculate the gradient: Calculate the gradient of the loss function relative to the parameter (compensation parameter) at the current point;

[0096] 3. Update parameters: Use the gradient descent formula to update parameters (compensation parameters);

[0097] Alternatively, the gradient descent formula can be expressed as:

[0098]

[0099] Among them, θ is the trained compensation parameter, α is the learning rate, is the gradient of the loss function.

[0100] 4. Repeat steps 2 and 3 above until the algorithm converges to the minimum value or reaches the predetermined number of iterations to obtain the final training model.

[0101] It should be noted that the specific implementation of the above-mentioned gradient descent algorithm is only for illustration and is not limited thereto.

[0102] The compensation parameter adjustment method described in the embodiment of the present invention is adopted, and data analysis is performed using multiple interface images of a large number of second display screens. The acquired quantization deviation values ​​and multiple compensation parameters are repeatedly trained to obtain an AI model with the quantization deviation value of the interface image compared to the corresponding standard interface as output and the Demura parameters for display compensation as output. The Demura parameters output by the AI ​​model can adjust the quantization deviation value of the interface image compared to the corresponding standard interface to the minimum value.

[0103] The compensation parameter adjustment method described in the embodiment of the present invention can deploy the obtained AI model in the processor of the display device after obtaining the AI ​​model based on the above-mentioned training process, and use the AI ​​model to obtain the Demura parameters for display compensation of the first display screen of the display device.

[0104] Combination Figure 3 As shown, the specific implementation process of adjusting the compensation parameters of the first display screen by using the method described in the embodiment of the present invention includes:

[0105] S301, providing a first display screen, wherein the first display screen is an AMOLD screen; optionally, controlling the first display screen to display an image with target display parameters, for example, displaying an image with a brightness of 500 nits and a grayscale of 32;

[0106] S302, taking a picture of the display interface of the first display screen using a camera to obtain an interface image of the first display screen;

[0107] Optionally, when the camera takes the first picture of the display interface of the first display screen, after step S302, the following steps S3021 to S3023 are executed, and the following steps S306 to S307 are executed simultaneously. Through steps S3021 to S3023, the interface image obtained by the camera is analyzed to obtain the mura data of the first display screen, and the obtained mura data is stored in the flash memory Flash, that is, steps S3021 to S3023 only need to be executed once; after the camera takes the second picture of the display interface of the first display screen, the following steps S3021 to S3023 are skipped, and only steps S303 to S307 are executed, that is, after each picture is taken to obtain the display interface of the first display screen, steps S303 to S307 can be repeatedly executed.

[0108] S3021, performing image processing on the interface image presented on the first display screen, optionally, the processing includes filtering and / or edge processing, to obtain a processed interface image;

[0109] S3022, performing data analysis on the processed interface image to obtain mura data of the first display screen;

[0110] S3023, storing the mura data in a memory; optionally, the mura data may be stored in a flash memory Flash by burning, wherein the mura data is used to record brightness information of each pixel on the first display screen.

[0111] S303, analyzing the interface image presented on the first display screen, and obtaining a quantized deviation value of the interface image relative to the corresponding standard interface; wherein the process of determining the quantized deviation value is as described above, and will not be repeated here;

[0112] S304, the quantization deviation value determined in step S303 is input into the AI ​​model obtained by the above training process, and the Demura parameter for display compensation is output;

[0113] S305, adjusting the display driving parameters of the first display screen according to the Demura parameters and the mura data of the first display screen obtained in step S3023, to obtain adjusted display parameters;

[0114] S306, adjusting the displayed image presented by the first display screen according to the adjusted display parameters, and obtaining a demura compensation effect according to the adjusted display image;

[0115] S307, returning to re-execute steps S301 to S307 until the obtained quantization deviation value is less than or equal to the preset deviation value;

[0116] That is, the interface image displayed on the first display screen is reacquired, data analysis is performed on the interface image to obtain the quantization deviation value, the quantization deviation value is input into the AI ​​model, the Demura parameter is obtained according to the output of the AI ​​model, the adjusted display parameter is obtained according to the Demura parameter and the mura data, and the interface image is adjusted according to the adjusted display parameter, until the obtained quantization deviation value is less than or equal to the preset deviation value.

[0117] However, it should be noted that the above steps S3021 to S3023 only need to be performed once.

[0118] Optionally, the preset deviation value is less than or equal to 0.05 and greater than or equal to 0.

[0119] By adopting the above implementation process, the quantitative deviation values ​​and corresponding compensation parameters of the display interfaces collected from multiple display interfaces of multiple display screens compared with the standard interface are trained and deduced to obtain a trained AI model. In the production process of the display panel, the Demura parameters of the display screen of the display panel are repeatedly adjusted using the AI ​​model until the quantitative deviation value of the interface image obtained is less than or equal to the preset deviation value. In this way, the adjustment of the Demura parameters of the display screen can be completed automatically without manual implementation, and each display screen in the mass production scale can be adjusted individually, thereby effectively ensuring the best display adjustment effect of each display screen.

[0120] In an embodiment of the present invention, optionally, a processor that executes the method described in the embodiment of the present invention may include multiple module parts, and the multiple module parts may include an AI chip (such as an application-specific integrated circuit ASIC), a display driver chip, and a processing chip, etc.; wherein the AI ​​chip is used to deploy an AI model; the display driver chip is used to input a display drive signal to the display screen to control the image display of the display screen; the processing chip is used to implement image processing of the interface image and timing control during the compensation parameter adjustment process, etc.

[0121] Based on the above-mentioned implementation structure, the AI ​​chip deploying the AI ​​model can be bound to the display device, and communicate with the Demura hardware IP module inside the display driver chip of the display device in real time to transmit the Demura parameters to the display driver chip. During the production process of the display device, the interface image of the display screen of the display device is obtained by camera shooting, and the interface image is analyzed, and the quantization bias value obtained by the data analysis is input into the AI ​​chip, and the Demura parameter is adjusted through the AI ​​model, and the adjustment process is cyclically executed according to the adjustment result of the Demura parameter until the quantization deviation value of the interface image obtained by shooting is less than or equal to the preset deviation value compared with the corresponding standard interface, then the adjustment process of the Demura parameter can be terminated, and it is determined that when the quantization deviation value is less than or equal to the preset deviation value, the corresponding quantization deviation value is input into the Demura parameter obtained in the AI ​​model, which is the final Demura parameter of the first display screen.

[0122] By adopting this implementation structure, an AI chip can be deployed on each display screen, which can realize differentiated adjustments during the mass production of display screens and further improve the compensation capability of the Demura algorithm.

[0123] Another aspect of the present invention is to provide a compensation parameter adjustment device. Figure 4 As shown, including:

[0124] The image acquisition module 410 is used to acquire an interface image displayed on the first display screen; the interface image shows a mura defect;

[0125] An analysis module 420 is used to perform data analysis on the interface image of the first display screen to obtain a quantitative deviation value of the interface image of the first display screen compared with a corresponding standard interface;

[0126] The processing module 430 is used to input the quantization deviation value into a pre-trained AI model, and obtain the Demura parameters for display compensation of the display screen according to the output of the AI ​​model; wherein the AI ​​model is obtained by training multiple acquired quantization deviation values ​​and multiple compensation parameters, the acquired quantization deviation value is the quantization deviation value between one of the interface images of the acquired multiple second display screens and the corresponding standard interface, and the compensation parameter is the parameter required to compensate for adjusting the acquired interface image to the corresponding standard interface.

[0127] By adopting the compensation parameter adjustment device described in the embodiment of the present invention, an AI model is obtained by training and deducing the quantized deviation value of the interface image of the display screen compared with the standard interface and the compensation parameter of the interface image compared with the standard interface, and the quantized deviation value of the collected interface image of the first display screen compared with the corresponding standard interface is analyzed to obtain the Demura parameters for display compensation of the display screen. In this way, by adopting this method, the adjustment of the Demura parameters can be completed automatically without manual implementation, and for each display screen in the mass production scale, individual adjustment can be achieved, thereby effectively ensuring the best display adjustment effect of each display screen.

[0128] Optionally, in the compensation parameter adjustment device, the processing module 430 is further used for:

[0129] According to the Demura parameters and the pre-obtained mura data of the first display screen, the display driving parameters of the first display screen are adjusted to obtain adjusted display parameters;

[0130] According to the adjusted display parameters, the interface image presented by the first display screen is adjusted.

[0131] Optionally, in the compensation parameter adjustment device, the processing module 430 is further used for:

[0132] Returning to the step of reacquiring the interface image displayed on the first display screen until the obtained quantization deviation value is less than or equal to a preset deviation value;

[0133] Determine that when the quantization deviation value is less than or equal to the preset deviation value, the corresponding quantization deviation value is input into the Demura parameter obtained by the AI ​​model, which is the final Demura parameter of the first display screen.

[0134] Optionally, in the compensation parameter adjustment device, the processing module 430 is further used for:

[0135] Performing filtering and / or edge processing on the interface image presented on the first display screen to obtain a processed interface image;

[0136] Obtaining mura data of the first display screen by performing data analysis on the processed interface image;

[0137] The mura data is stored in a memory.

[0138] Optionally, in the compensation parameter adjustment device, the processing module 430 is further used for:

[0139] Performing data analysis on the obtained multiple interface images of the multiple second display screens to obtain a quantitative deviation value of each interface image of the second display screen compared with a corresponding standard interface;

[0140] Calculating compensation data corresponding to the interface image of the second display screen according to each interface image of the plurality of second display screens collected and the corresponding standard interface;

[0141] Using a principal component analysis (PCA) learning algorithm to determine basic compensation data from a plurality of compensation data;

[0142] The quantization deviation value corresponding to the interface images of the multiple second display screens is set as the acquisition quantization deviation value, and the acquisition quantization deviation value is used as the input of the model training, and the basic compensation data is used as the output of the model training, and the gradient descent algorithm is used for training to obtain the AI ​​model.

[0143] Optionally, in the compensation parameter adjustment device, the analysis module 420 performs data analysis on the interface image to obtain a quantitative deviation value of the interface image compared to a corresponding standard interface, including:

[0144] Performing data analysis on the interface image to obtain an effect quantification value corresponding to the interface image; the effect quantification value is a characterization value used to indicate the severity of the mura defect;

[0145] The effect quantization value corresponding to the interface image is compared with the effect quantization value corresponding to the standard interface to obtain the quantization deviation value.

[0146] Optionally, in the compensation parameter adjustment device, the analysis module 420 performs data analysis on the interface image to obtain a quantitative effect value corresponding to the interface image, including:

[0147] After dividing the interface image into a plurality of sub-regions, the number of mura defects in each sub-region is sequentially obtained using an image processing algorithm;

[0148] An effect quantification value corresponding to the interface image is determined according to the total number of mura defects in the multiple sub-regions and the number of the multiple sub-regions.

[0149] Optionally, in the compensation parameter adjustment device, the analysis module 420 is further used for:

[0150] Performing simulation processing on the interface image to obtain a standard interface corresponding to the interface image;

[0151] Set the effect quantization value corresponding to the standard interface to the preset value.

[0152] The specific implementation of the compensation parameter adjustment device described in the embodiment of the present invention can be combined with the detailed description of the compensation parameter adjustment method, and will not be repeated here. Another embodiment of the present invention also provides a display device, including a processor, a memory, and a program stored in the memory and executable on the processor, wherein the program is executed by the processor to implement the compensation parameter adjustment method described in any one of the above items.

[0153] For the specific implementation of the compensation parameter adjustment method implemented by the processor in the display device according to the embodiment of the present invention, please refer to the above detailed description, which will not be repeated here.

[0154] One embodiment of the present invention further provides a computer program product, which includes computer instructions, and when the computer instructions are executed by a processor, the steps in the compensation parameter adjustment method described in any one of the above are implemented. Optionally, the embodiment of the present invention can take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program codes.

[0155] The computer program product described in the embodiment of the present invention includes computer instructions that implement the various processes of the authentication method embodiment shown above when executed by a processor, and can achieve the same technical effect. To avoid repetition, they are not described here.

[0156] The above is a preferred embodiment of the present invention. It should be pointed out that for ordinary personnel in this technical field, several improvements and modifications can be made without departing from the principles of the present invention. These improvements and modifications should also be regarded as the scope of protection of the present invention.

Claims

1. A compensation parameter adjustment method, characterized in that: include: Acquire the interface image displayed on the first display screen; The interface image shows a mura defect; Performing data analysis on the interface image of the first display screen to obtain a quantitative deviation value of the interface image of the first display screen compared with a corresponding standard interface; The quantization deviation value is input into a pre-trained AI model, and the Demura parameters for display compensation of the first display screen are obtained according to the output of the AI ​​model; wherein the AI ​​model is obtained by training multiple acquired quantization deviation values ​​and multiple compensation parameters, the acquired quantization deviation value is the quantization deviation value between the acquired interface images of the multiple second display screens and the corresponding standard interface, and the compensation parameter is the parameter required to compensate for adjusting the acquired interface image to the corresponding standard interface.

2. The compensation parameter adjustment method according to claim 1, characterized in that: The method further comprises: According to the Demura parameters and the pre-obtained mura data of the first display screen, the display driving parameters of the first display screen are adjusted to obtain adjusted display parameters; According to the adjusted display parameters, the interface image presented by the first display screen is adjusted.

3. The compensation parameter adjustment method according to claim 2, characterized in that: The method further comprises: Returning to the step of reacquiring the interface image displayed on the first display screen until the obtained quantization deviation value is less than or equal to a preset deviation value; Determine that when the quantization deviation value is less than or equal to the preset deviation value, the corresponding quantization deviation value is input into the Demura parameter obtained by the AI ​​model, which is the final Demura parameter of the first display screen.

4. The compensation parameter adjustment method according to claim 2, characterized in that: The method further comprises: Performing filtering and / or edge processing on the interface image presented on the first display screen to obtain a processed interface image; Obtaining mura data of the first display screen by performing data analysis on the processed interface image; The mura data is stored in a memory.

5. The compensation parameter adjustment method according to claim 1, characterized in that: The method further comprises: Performing data analysis on the obtained multiple interface images of the multiple second display screens to obtain a quantitative deviation value of each interface image of the second display screen compared with a corresponding standard interface; Calculating compensation data corresponding to the interface image of the second display screen according to each interface image of the plurality of second display screens collected and the corresponding standard interface; Using a principal component analysis (PCA) learning algorithm to determine basic compensation data from a plurality of compensation data; The quantization deviation value corresponding to the interface images of the multiple second display screens is set as the acquisition quantization deviation value, and the acquisition quantization deviation value is used as the input of the model training, and the basic compensation data is used as the output of the model training, and the gradient descent algorithm is used for training to obtain the AI ​​model.

6. The compensation parameter adjustment method according to claim 1 or 5, characterized in that: Performing data analysis on the interface image to obtain a quantitative deviation value of the interface image compared with a corresponding standard interface includes: Performing data analysis on the interface image to obtain an effect quantification value corresponding to the interface image; the effect quantification value is a characterization value used to indicate the severity of the mura defect; The effect quantization value corresponding to the interface image is compared with the effect quantization value corresponding to the standard interface to obtain the quantization deviation value.

7. The compensation parameter adjustment method according to claim 6, characterized in that: Performing data analysis on the interface image to obtain a quantitative value of an effect corresponding to the interface image includes: After dividing the interface image into a plurality of sub-regions, the number of mura defects in each sub-region is sequentially obtained using an image processing algorithm; Determining a defect characterization value corresponding to the interface image according to the total number of mura defects in the multiple sub-regions and the number of the multiple sub-regions; According to the defect characterization value, the corresponding effect quantification value is determined.

8. The compensation parameter adjustment method according to claim 6, characterized in that: The method further comprises: Performing simulation processing on the interface image to obtain a standard interface corresponding to the interface image; Set the effect quantization value corresponding to the standard interface to the preset value.

9. A compensation parameter adjustment device, characterized in that: include: An image acquisition module, used to acquire an interface image displayed on the first display screen; The interface image shows a mura defect; An analysis module, configured to perform data analysis on the interface image of the first display screen to obtain a quantitative deviation value of the interface image of the first display screen compared with a corresponding standard interface; A processing module is used to input the quantization deviation value into a pre-trained AI model, and obtain Demura parameters for display compensation of the display screen according to the output of the AI ​​model; wherein the AI ​​model is obtained by training multiple acquired quantization deviation values ​​and multiple compensation parameters, the acquired quantization deviation value is the quantization deviation value between one of the interface images of the acquired multiple second display screens and the corresponding standard interface, and the compensation parameter is the parameter required to compensate for adjusting the acquired interface image to the corresponding standard interface.

10. A display device, characterized in that: The method comprises a processor, a memory and a program stored in the memory and executable on the processor, wherein when the program is executed by the processor, the compensation parameter adjustment method according to any one of claims 1 to 8 is implemented.

11. A computer program product, characterized in that The method comprises computer instructions, which, when executed by a processor, implement the steps in the compensation parameter adjustment method according to any one of claims 1 to 8.

Images