Display uniformity compensation method based on machine learning, storage medium and terminal

Abstract

The invention discloses a display uniformity compensation method based on machine learning, a storage medium and a terminal. The method comprises the following steps: providing a plurality of target gray scale data; acquiring a plurality of acquisition pictures; acquiring compensation result gray scale data of each acquired picture; providing a machine learning model; completing training and testing of a machine learning model based on the target gray scale data and the compensation result gray scale data; solidifying the model architecture and the model parameters of the machine learning model to the to-be-compensated display equipment; and outputting the application gray scale data of the to-be-compensated display equipment based on the machine learning model. Based on the machine learning model, the application gray scale data of the to-be-compensated display equipment can be quickly output, and the display uniformity compensation effect and efficiency of the to-be-compensated display equipment are effectively improved. In addition, the model architecture of the machine learning model passing the test and the model parameters are solidified to the to-be-compensated display device, compensation data can be effectively reduced or even completely not required to be stored any more, and the cost of data storage and driving operation is greatly reduced.

Description

Technical Field

[0001] The present invention relates to the field of display technology, and in particular to a display uniformity compensation method, storage medium and terminal based on machine learning. Background Art

[0002] Display uniformity compensation technology is a digital processing technology that improves the uneven optical characteristics of flat panel display products. It adjusts the brightness of sub-pixels by controlling grayscale data, so that pixels with large brightness or chromaticity deviations can work according to the expected optical goals.

[0003] The Demura technology used in active matrix organic light emitting diode (AMOLED) products is a practical application of display uniformity compensation technology. It performs fitting calculations by analyzing the relationship between sub-pixel brightness and grayscale, and then calculates the compensation amount for each pixel. The compensation data is stored in the Flash of the display module, and the driver IC calls the compensation data in real time to compensate the input display image.

[0004] Micro Light Emitting Diode Display (MicroLED) as a new display technology also requires display uniformity compensation technology for optical improvement. Due to the characteristics of process and materials, the industry generally believes that MicroLED needs to compensate for uniformity of brightness and color at the same time.

[0005] However, there are still many problems in the prior art when MicroLED is used to compensate for display uniformity. Summary of the invention

[0006] The technical problem solved by the present invention is to provide a display uniformity compensation method, storage medium and terminal based on machine learning, so as to improve the display uniformity compensation effect and efficiency, and reduce the cost of data storage and driving operation.

[0007] In order to solve the above problems, the present invention provides a display uniformity compensation method based on machine learning, including: providing a number of target grayscale data; obtaining a number of corresponding acquisition pictures based on the target grayscale data, and the acquisition pictures are divided into a training set and a test set; obtaining compensation result grayscale data of each acquisition picture based on data calculation; providing a machine learning model; inputting the target grayscale data and the compensation result grayscale data of each acquisition picture in the training set into the machine learning model for model training; after the model training, inputting the target grayscale data of each acquisition picture in the test set into the trained machine learning model for model training. Test, based on the corresponding predicted grayscale data output by the machine learning model; compare each of the predicted grayscale data with the compensation result grayscale data of the corresponding captured image in the test set until the test accuracy reaches the preset accuracy, and complete the model test of the machine learning model; solidify the model architecture of the machine learning model that passes the test to the display device to be compensated; write the model parameters of the machine learning model that passes the test to the display device to be compensated; use the machine learning model to calculate the target grayscale data to be compensated that is input into the display device to be compensated, and output the application grayscale data of the display device to be compensated based on the machine learning model.

[0008] Optionally, the method for obtaining a number of corresponding captured images based on a number of the target grayscale data includes: providing a test display device; inputting a number of the target grayscale data into the test display device, and displaying a number of corresponding display images based on the test display device; and using a picture capture device to shoot each of the display images to obtain the corresponding captured images.

[0009] Optionally, the test display device and the display device to be compensated are the same display device, or display devices from the same process batch.

[0010] Optionally, the method for obtaining compensation result grayscale data of each captured image based on data calculation includes: extracting measurement display data of each pixel point on the captured image; providing target display data of each pixel point on the captured image; obtaining a correlation function between the target grayscale data and the target display data of the captured image based on data fitting; adjusting the target grayscale data based on the correlation function to adjust the measurement display data until the difference between the adjusted measurement display data and the target display data meets a threshold, and using the adjusted target grayscale data as the compensation result grayscale data.

[0011] Optionally, the measured display data includes: measured brightness data and measured chromaticity data; and the target display data includes: target brightness data and target chromaticity data.

[0012] Optionally, each of the predicted grayscale data is compared with the compensation result grayscale data of the corresponding captured image in the test set, and whether the predicted grayscale data passes the test is determined based on an evaluation function.

[0013] Optionally, the method for determining whether the predicted grayscale data passes the test based on an evaluation function includes: inputting the predicted grayscale data and the corresponding compensation result grayscale data into the evaluation function to obtain an evaluation value; when the evaluation value is within a preset evaluation threshold range, determining that the predicted grayscale data passes the test.

[0014] Optionally, the evaluation function includes: a root mean square error function and a goodness of fit function.

[0015] Optionally, the method of training until the test accuracy reaches a preset accuracy includes: when the test accuracy of the machine learning model does not reach the preset accuracy, increasing the number of model training of the machine learning model until the test accuracy of the machine learning model reaches the preset accuracy.

[0016] Optionally, the method for increasing the number of model training times of the machine learning model includes: inputting the target grayscale data and the compensation result grayscale data of the captured image in the test set that failed the test during the test into the machine learning model for model training.

[0017] Optionally, the number of the captured images in the training set is greater than the number of the captured images in the test set.

[0018] Optionally, the machine learning model includes: a support vector machine model, a neural network model or a random forest model.

[0019] Optionally, the acquisition screen is divided into a plurality of display areas, the pixels on each display area display the same color, and the pixels on different display areas display different colors.

[0020] Optionally, the colors displayed by each pixel on the acquisition screen are any combination.

[0021] Optionally, the ratio of the number of pixels displaying a single red color on the acquisition screen to the number of all pixels on the acquisition screen is greater than 50%.

[0022] Optionally, the ratio of the number of pixels displaying a single green color on the acquisition picture to the number of all pixels on the acquisition picture is less than 20%.

[0023] Optionally, the ratio of the number of pixels displaying a single blue color on the acquisition picture to the number of all pixels on the acquisition picture is less than 20%.

[0024] Optionally, the ratio of the number of pixels displaying mixed colors on the acquisition screen to the number of all pixels on the acquisition screen is 10% to 30%.

[0025] Correspondingly, the technical solution of the present invention also provides a storage medium on which computer instructions are stored, and when the computer instructions are executed, the steps of the method described in any one of the above technical solutions are executed.

[0026] Correspondingly, the technical solution of the present invention also provides a terminal, including a memory and a processor, wherein the memory stores computer instructions that can be executed on the processor, and when the processor executes the computer instructions, the steps of the method described in any one of the above technical solutions are executed.

[0027] Compared with the prior art, the technical solution of the present invention has the following advantages:

[0028] In the display uniformity compensation method based on machine learning of the technical solution of the present invention, the machine learning model is trained and tested by the target grayscale data and the compensation result grayscale data of the several captured images. After the machine learning model passes the test, the application grayscale data of the display device to be compensated can be quickly output based on the machine learning model, thereby effectively improving the display uniformity compensation effect and efficiency of the display device to be compensated. In addition, the model architecture and model parameters of the machine learning model that has passed the test are solidified to the display device to be compensated, which can effectively reduce or even completely eliminate the need to store compensation data, greatly reducing the cost of data storage and drive calculation.

[0029] Furthermore, the acquisition screen is divided into several display areas, the pixels on each display area display the same color, and the pixels on different display areas display different colors; or the colors displayed by each pixel on the acquisition screen are any combination. The colors displayed by each pixel on each acquisition screen are kept different as much as possible, so that the machine learning model has more comprehensive data support, thereby improving the accuracy of the application grayscale data output by the machine learning model, and further improving the display uniformity compensation effect of the display device to be compensated.

[0030] Furthermore, the ratio of the number of pixels showing a single red color on the acquisition screen to the number of all pixels on the acquisition screen is greater than 50%. Since the target grayscale data corresponding to the red color and the measured display data have relatively large regular fluctuations, it is necessary to set as many red color ratios as possible so that the machine learning model can be trained on more data corresponding to the red color, thereby improving the accuracy of the machine learning model. BRIEF DESCRIPTION OF THE DRAWINGS

[0031] Figure 1 This is a schematic diagram of the relationship between grayscale and brightness in AMOLED;

[0032] Figure 2 This is a schematic diagram of the relationship between grayscale, brightness and chromaticity in MicroLED;

[0033] Figure 3 is a flow chart of a display uniformity compensation method based on machine learning according to an embodiment of the present invention;

[0034] Figure 4 It is a schematic diagram of a machine learning model in a display uniformity compensation method based on machine learning in an embodiment of the present invention. DETAILED DESCRIPTION

[0035] In order to make the above-mentioned purposes, features and advantages of the present invention more obvious and easy to understand, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0036] In the description of the present invention, it should be understood that the terms "upper", "lower", "top surface", "bottom surface" and the like indicate positions or positional relationships based on the positions or positional relationships shown in the drawings, and are only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the positions or elements referred to must have a specific orientation, be constructed and operate in a specific orientation, and therefore cannot be understood as limitations of the present invention. In addition, the terms "first" and "second" are used for descriptive purposes only and cannot be understood as indicating or implying relative importance.

[0037] As described in the background art, there are still many problems in the prior art when MicroLED is used to compensate for display uniformity, which will be described in detail below with reference to the accompanying drawings.

[0038] Figure 1 This is a schematic diagram of the relationship between grayscale and brightness in AMOLED; Figure 2 Schematic diagram of the relationship between grayscale, brightness and chromaticity in MicroLED.

[0039] In AMOLED (Active-matrix organic light-emitting diode) products, due to the high accuracy of color, when performing display uniformity compensation, usually only the brightness of the AMOLED product is compensated. Moreover, the grayscale (R, G, B) and brightness (Lr, Lg, Lb) of AMOLED products are in a one-to-one relationship and do not interfere with each other (such as Figure 1 shown).

[0040] However, in MicroLED (Micro Light Emitting Diode Display) products, it is necessary to compensate for both the chromaticity and brightness of MicroLED products. In addition, there is a many-to-many cross-correlation between the grayscale (R, G, B) and brightness (Lr, Lg, Lb) and chromaticity ((Xw, Yw), (Xr, Yr), (Xg, Yg), (Xb, Yb)) of MicroLED products, which interfere with each other and make it difficult to find regularities in the data (such as Figure 2 shown).

[0041] The applicant has found that the reason for the above problems is that the luminous characteristics of MicroLED products are complex and there are too many influencing factors. This is mainly because the device efficiency and luminous wavelength of MicroLED will change with the change of current density. The relevant change rules make it very difficult to find and construct a functional relationship that can accurately describe it. The method of cyclic iteration will increase the amount of calculation. The brightness and chromaticity of the display screen vary greatly at different brightness levels, which also makes it impossible to use a set of compensation data to achieve full grayscale compensation, resulting in a surge in data storage and the inability to use a simple optical model to build compensation logic. If the traditional optical algorithm is used to calculate the results, the required amount of calculation and compensation will increase by more than 2 orders of magnitude compared to AMOLED products. If the compensation data is stored in the Flash of the display module, and then the compensation data is called in real time by the driver IC to compensate the input display screen, it will put forward high requirements on the storage capacity of the Flash and the computing power of the driver IC, which directly affects the feasibility of mass production.

[0042] The applicant has further discovered that a machine learning algorithm is an algorithm that learns patterns and rules from data to make predictions and decisions. It can better fit nonlinear relationships, effectively process large-scale data, improve the scalability and efficiency of the algorithm, and better adapt to unseen data, rather than just the ability to fit training data. Therefore, compensating for the display uniformity of MicroLED products based on machine learning can effectively improve the display uniformity compensation effect and efficiency.

[0043] Figure 3 : is a flow chart of a display uniformity compensation method based on machine learning according to an embodiment of the present invention, comprising:

[0044] Step S101, providing some target grayscale data;

[0045] Step S102, acquiring a number of corresponding acquisition images based on the target grayscale data, wherein the acquisition images are divided into a training set and a test set;

[0046] Step S103, obtaining compensation result grayscale data of each captured image based on data calculation;

[0047] Step S104, providing a machine learning model;

[0048] Step S105, inputting the target grayscale data and the compensation result grayscale data of each captured image in the training set into the machine learning model for model training;

[0049] Step S106, after model training, inputting the target grayscale data of each captured image in the test set into the trained machine learning model for model testing, and outputting corresponding predicted grayscale data based on the machine learning model;

[0050] Step S107, comparing each of the predicted grayscale data with the compensation result grayscale data of the corresponding captured image in the test set until the test accuracy reaches a preset accuracy, thereby completing the model test of the machine learning model;

[0051] Step S108, solidifying the model architecture of the machine learning model that has passed the test to the display device to be compensated;

[0052] Step S109, writing the model parameters of the machine learning model that has passed the test into the display device to be compensated;

[0053] In step S110 , the target grayscale data to be compensated input into the display device to be compensated is calculated by the machine learning model, and application grayscale data of the display device to be compensated is output based on the machine learning model.

[0054] The steps of the display uniformity compensation method based on machine learning will be described in detail below with reference to the accompanying drawings.

[0055] Figure 4 It is a schematic diagram of a machine learning model in a display uniformity compensation method based on machine learning in an embodiment of the present invention.

[0056] In step S101 , a plurality of target grayscale data are provided.

[0057] It should be noted that, in this embodiment, some of the target grayscale data may be input into the test display device through manual setting.

[0058] In step S102, a number of corresponding acquisition pictures are acquired based on a number of the target grayscale data, and the number of the acquisition pictures are divided into a training set and a test set.

[0059] In this embodiment, the method for obtaining a number of corresponding captured images based on a number of the target grayscale data includes: providing a test display device; inputting a number of the target grayscale data into the test display device, and displaying a number of corresponding display images based on the test display device; and using a picture capture device to shoot each of the display images to obtain the corresponding captured images.

[0060] It should be noted that, in this embodiment, the target grayscale data does not include brightness data and chromaticity data, and the specific form is RxxxGxxxBxxx (where xxx is 000-255). Only the captured image can obtain the brightness data and chromaticity data of each pixel.

[0061] In this embodiment, the number of the captured images in the training set is greater than the number of the captured images in the test set.

[0062] In step S103, the compensation result grayscale data of each captured image is obtained based on data calculation.

[0063] In this embodiment, the method for obtaining compensation result grayscale data of each captured image based on data calculation includes: extracting measurement display data of each pixel point on the captured image; providing target display data of each pixel point on the captured image; obtaining a correlation function between the target grayscale data and the target display data of the captured image based on data fitting; adjusting the target grayscale data based on the correlation function to adjust the measurement display data until the difference between the adjusted measurement display data and the target display data meets a threshold, and using the adjusted target grayscale data as the compensation result grayscale data.

[0064] In this embodiment, the measured display data includes: measured brightness data and measured chromaticity data; the target display data includes: target brightness data and target chromaticity data.

[0065] It should be noted that in this embodiment, when each pixel has no process deviation, that is, there is no need for display uniformity compensation, the target brightness data and the target chromaticity data can be calculated based on the input target grayscale data, that is, the target brightness data and the target chromaticity data are brightness data and chromaticity data under ideal conditions, and the data are known.

[0066] Data fitting is used to determine the specific compensation amount for the target grayscale data. For example, if the measured brightness data is found to be 10 nits higher than the target brightness data through shooting, the brightness difference here can be converted into grayscale data difference through data fitting to achieve compensation for the display data.

[0067] In a specific embodiment, if it is found that the measured brightness data of a certain pixel point when displaying the target grayscale data as R160G000B000 is 10nits higher than the target brightness data, through data fitting it can be obtained that the target brightness data can be achieved by reducing the red color in the target grayscale data by 2 grayscales, then the compensated target grayscale data becomes R158G000B000, and R158G000B000 is the compensated result grayscale data, and the compensated result grayscale data is known and correct.

[0068] In step S104, a machine learning model is provided.

[0069] In this embodiment, the machine learning model includes: a support vector machine model, a neural network model or a random forest model.

[0070] In this embodiment, the acquisition screen is divided into several display areas, and the pixels on each display area display the same color, and the pixels on different display areas display different colors; or the colors displayed by each pixel on the acquisition screen are any combination. The colors displayed by each pixel on each acquisition screen are kept different as much as possible, so that the machine learning model has more comprehensive data support, thereby improving the accuracy of the application grayscale data output by the machine learning model, and further improving the display uniformity compensation effect of the display device to be compensated.

[0071] In this embodiment, the ratio of the number of pixels showing a single red color on the acquisition screen to the number of all pixels on the acquisition screen is greater than 50%. Optionally, the ratio of the number of pixels showing a single red color on the acquisition screen to the number of all pixels on the acquisition screen is 70% or 80%.

[0072] Since micro-LEDs generally include red micro-LEDs, green micro-LEDs, and blue micro-LEDs, and the luminous efficiency of red micro-LEDs is lower than that of green micro-LEDs or blue micro-LEDs, the target grayscale data of red micro-LEDs corresponding to red colors differ or fluctuate greatly. Since the target grayscale data corresponding to the red color and the measured display data fluctuate greatly, in order to avoid large deviations in the red color acquisition data, it is necessary to set as many red color ratios as possible so that the machine learning model can fully train the data corresponding to the red color, thereby improving the accuracy of the machine learning model.

[0073] In this embodiment, the ratio of the number of pixels showing a single green color on the acquisition screen to the number of all pixels on the acquisition screen is less than 20%. Optionally, the ratio of the number of pixels showing a single green color on the acquisition screen to the number of all pixels on the acquisition screen is 10% or 5%. Since micro-LEDs generally include red micro-LEDs, green micro-LEDs and blue micro-LEDs, the luminous efficiency of green micro-LEDs is higher than that of red micro-LEDs, and the stability of the green color corresponding to the green micro-LEDs is better than that of the red color, and the volatility of the acquisition data is small. In order to collect more data on other colors, the ratio of the number of pixels showing a single green color to the number of all pixels on the acquisition screen is less than 20%, so that the machine learning model can fully train the data corresponding to other colors, thereby improving the accuracy of the machine learning model.

[0074] In this embodiment, the ratio of the number of pixels showing a single blue color on the acquisition screen to the number of all pixels on the acquisition screen is less than 20%. Optionally, the ratio of the number of pixels showing a single blue color on the acquisition screen to the number of all pixels on the acquisition screen is 10% or 5%. Since micro-LEDs generally include red micro-LEDs, green micro-LEDs and blue micro-LEDs, the luminous efficiency of blue micro-LEDs is higher than that of red micro-LEDs, and the stability of the blue color corresponding to the blue micro-LEDs is better than that of the red color, and the volatility of the acquisition data is small. In order to collect more data for other colors, the ratio of the number of pixels showing a single blue color to the number of all pixels on the acquisition screen is less than 20%, so that the machine learning model can fully train the data corresponding to other colors, thereby improving the accuracy of the machine learning model.

[0075] Optionally, the ratio of the number of pixels of a single red color to the number of all pixels on the acquisition screen is 80%, the ratio of the number of pixels of a single blue color to the number of all pixels on the acquisition screen is 10%, and the ratio of the number of pixels of a single blue color to the number of all pixels on the acquisition screen is 10%.

[0076] In this embodiment, the ratio of the number of pixels displaying mixed colors on the acquisition screen to the number of all pixels on the acquisition screen is 10% to 30%. Optionally, the ratio of the number of pixels displaying mixed colors on the acquisition screen to the number of all pixels on the acquisition screen is 20%.

[0077] In step S105, the target grayscale data and the compensation result grayscale data of each captured image in the training set are input into the machine learning model for model training.

[0078] In this embodiment, during the model training stage, the target grayscale data of the captured image is taken as a condition, and the compensation result grayscale data of the captured image is taken as a result, and both are simultaneously input into the machine learning model, and the machine learning model learns and establishes a mapping relationship by itself.

[0079] In step S106, after model training, the target grayscale data of each captured image in the test set is input into the trained machine learning model for model testing, and corresponding predicted grayscale data is output based on the machine learning model.

[0080] In this embodiment, during the model testing phase, only the target grayscale data of the captured image is input as a condition into the machine learning model, and the machine learning model learns and establishes a mapping relationship based on the training phase, and outputs the corresponding predicted grayscale data. However, whether the predicted grayscale data is accurate still needs to be verified based on the known and correct compensation result grayscale data.

[0081] In step S107, each of the predicted grayscale data is compared with the compensation result grayscale data of the corresponding acquisition picture in the test set until the test accuracy reaches a preset accuracy, thereby completing the model test of the machine learning model.

[0082] In this embodiment, each of the predicted grayscale data is compared with the compensation result grayscale data of the corresponding captured image in the test set, and it is determined whether the predicted grayscale data passes the test based on an evaluation function.

[0083] In this embodiment, the method for judging whether the predicted grayscale data passes the test based on the evaluation function includes: inputting the predicted grayscale data and the corresponding compensation result grayscale data into the evaluation function to obtain an evaluation value; when the evaluation value is within a preset evaluation threshold range, judging that the predicted grayscale data passes the test.

[0084] In this embodiment, the evaluation function includes: a root mean square error function (RMSE) and a goodness of fit function (R-squared, R2). The root mean square error function and the goodness of fit function are used to measure the index of the difference between the predicted parameter value and the actual parameter value. When the difference index between the two is within the preset evaluation threshold range, it is considered that the two match.

[0085] For example, in a specific embodiment, if the output predicted grayscale data is R170G083B000, and the compensation result grayscale data is R170G085B000, although the two are not exactly the same, the evaluation value obtained by calculation based on the evaluation function is within the preset evaluation threshold range, then the predicted grayscale data is considered to have passed the test.

[0086] In this embodiment, the method of training until the test accuracy reaches the preset accuracy includes: when the test accuracy of the machine learning model does not reach the preset accuracy, increasing the number of model training of the machine learning model until the test accuracy of the machine learning model reaches the preset accuracy.

[0087] In this embodiment, the method for increasing the number of model training times of the machine learning model includes: inputting the target grayscale data and the compensation result grayscale data of the captured image in the test set that failed the test during the test into the machine learning model for model training.

[0088] In step S108, the model architecture of the machine learning model that has passed the test is solidified to the display device to be compensated.

[0089] In this embodiment, the model architecture is solidified into the hardware deployment of the algorithm in the display device to be compensated. The general process is to first use the FPGA platform to write the model architecture in the form of an algorithm, and perform functional simulation verification. After the verification, the algorithm is converted into RTL or netlist and handed over to the chip manufacturer for design integration and finally tape-out.

[0090] In this embodiment, the test display device and the display device to be compensated are the same display device. Since the display effects of different display devices may be different, that is, the display uniformity compensation of each display device is different. Therefore, in order to avoid display differences between different display devices and achieve the best display uniformity compensation effect, the test display device and the display device to be compensated are the same display device, that is, each display device to be compensated is debugged for display uniformity compensation before leaving the factory.

[0091] In other embodiments, if the process technology can ensure that the display differences of display devices of the same process batch are small enough, the test display device and the display device to be compensated can also be display devices of the same process batch, and only one or several display devices need to be debugged, and the debugging results can be applied to other display devices of the same process batch, thereby avoiding the step of debugging the display uniformity compensation for each display device to be compensated before leaving the factory, thereby improving production efficiency.

[0092] In step S109, the model parameters of the machine learning model that has passed the test are written into the display device to be compensated.

[0093] In this embodiment, the model parameters are fixed in the driver IC or TCON through One Time Programmable (OTP) or other data writing forms.

[0094] In step S110, the target grayscale data to be compensated input into the display device to be compensated is calculated by the machine learning model, and application grayscale data of the display device to be compensated is output based on the machine learning model.

[0095] The machine learning model is trained and tested through the target grayscale data and the compensation result grayscale data of several captured images. After the machine learning model passes the test, the application grayscale data of the display device to be compensated can be quickly output based on the machine learning model, thereby effectively improving the display uniformity compensation effect and efficiency of the display device to be compensated. In addition, the model architecture and model parameters of the machine learning model that has passed the test are solidified to the display device to be compensated, which can effectively reduce or even completely eliminate the need to store compensation data, greatly reducing the cost of data storage and drive calculations.

[0096] Correspondingly, an embodiment of the present invention further provides a storage medium on which computer instructions are stored. When the computer instructions are executed, the steps of any one of the above-mentioned embodiment methods are executed.

[0097] Correspondingly, an embodiment of the present invention further provides a terminal, including a memory and a processor, wherein the memory stores computer instructions that can be executed on the processor, and when the processor executes the computer instructions, the steps of any one of the above-mentioned embodiment methods are executed.

[0098] Although the present invention is disclosed as above, the present invention is not limited thereto. Any person skilled in the art can make various changes and modifications without departing from the spirit and scope of the present invention. Therefore, the protection scope of the present invention shall be subject to the scope defined by the claims.

Claims

1. A display uniformity compensation method based on machine learning, characterized in that: include: Provide some target grayscale data; Acquire a number of corresponding acquisition images based on the target grayscale data, wherein the acquisition images are divided into a training set and a test set; Obtaining compensation result grayscale data of each captured image based on data calculation; Provide machine learning models; Inputting the target grayscale data and the compensation result grayscale data of each captured image in the training set into the machine learning model for model training; After model training, the target grayscale data of each captured image in the test set is input into the trained machine learning model for model testing, and corresponding predicted grayscale data is output based on the machine learning model; Compare each of the predicted grayscale data with the compensation result grayscale data of the corresponding captured image in the test set until the test accuracy reaches a preset accuracy, thereby completing the model test of the machine learning model; Solidify the model architecture of the machine learning model that has passed the test to the display device to be compensated; The model parameters of the tested machine learning model are written into the display device to be compensated; the target grayscale data to be compensated input into the display device to be compensated is calculated by the machine learning model, and the application grayscale data of the display device to be compensated is output based on the machine learning model.

2. The display uniformity compensation method based on machine learning according to claim 1, characterized in that: The method for obtaining a number of corresponding acquisition images based on a number of the target grayscale data includes: providing a test display device; inputting a number of the target grayscale data into the test display device, and displaying a number of corresponding display images based on the test display device; and using a picture acquisition device to shoot each of the display images to obtain the corresponding acquisition images.

3. The display uniformity compensation method based on machine learning according to claim 2, characterized in that: The test display device and the display device to be compensated are the same display device, or display devices from the same process batch.

4. The display uniformity compensation method based on machine learning according to claim 1, characterized in that: The method for obtaining compensation result grayscale data of each collected picture based on data calculation includes: extracting measurement display data of each pixel point on the collected picture; providing target display data of each pixel point on the collected picture; obtaining the correlation function between the target grayscale data and the target display data of the collected picture based on data fitting; adjusting the target grayscale data based on the correlation function to adjust the measurement display data until the difference between the adjusted measurement display data and the target display data meets the threshold, and using the adjusted target grayscale data as the compensation result grayscale data.

5. The display uniformity compensation method based on machine learning as claimed in claim 4, characterized in that: The measured display data includes: measured brightness data and measured chromaticity data; the target display data includes: target brightness data and target chromaticity data.

6. The display uniformity compensation method based on machine learning according to claim 1, characterized in that: Each of the predicted grayscale data is compared with the compensation result grayscale data of the corresponding acquisition picture in the test set, and it is determined whether the predicted grayscale data passes the test based on an evaluation function.

7. The display uniformity compensation method based on machine learning according to claim 6, characterized in that: The method for judging whether the predicted grayscale data passes the test based on the evaluation function includes: inputting the predicted grayscale data and the corresponding compensation result grayscale data into the evaluation function to obtain an evaluation value; when the evaluation value is within a preset evaluation threshold range, judging that the predicted grayscale data passes the test.

8. The display uniformity compensation method based on machine learning according to claim 6, characterized in that: The evaluation function includes: a root mean square error function and a goodness of fit function.

9. The display uniformity compensation method based on machine learning according to claim 1, characterized in that: The method for training the machine learning model until the test accuracy reaches the preset accuracy includes: when the test accuracy of the machine learning model does not reach the preset accuracy, increasing the number of model training times of the machine learning model until the test accuracy of the machine learning model reaches the preset accuracy.

10. The display uniformity compensation method based on machine learning according to claim 9, characterized in that: The method for increasing the number of model training times of the machine learning model includes: inputting the target grayscale data and the compensation result grayscale data of the captured image in the test set that failed the test during the test into the machine learning model for model training.

11. The display uniformity compensation method based on machine learning according to claim 1, characterized in that: The number of the captured images in the training set is greater than the number of the captured images in the test set.

12. The display uniformity compensation method based on machine learning according to claim 1, characterized in that: The machine learning model includes: a support vector machine model, a neural network model or a random forest model.

13. The display uniformity compensation method based on machine learning according to claim 1, characterized in that: The acquisition screen is divided into a plurality of display areas, the pixels on each display area display the same color, and the pixels on different display areas display different colors.

14. The display uniformity compensation method based on machine learning according to claim 1, characterized in that: The colors displayed by each pixel on the acquisition screen are any combination.

15. The display uniformity compensation method based on machine learning according to claim 13 or 14, characterized in that: The ratio of the number of pixels displaying a single red color on the acquisition picture to the number of all pixels on the acquisition picture is greater than 50%.

16. The display uniformity compensation method based on machine learning according to claim 13 or 14, characterized in that: The ratio of the number of pixels displaying a single green color on the acquisition picture to the number of all pixels on the acquisition picture is less than 20%.

17. The display uniformity compensation method based on machine learning according to claim 13 or 14, characterized in that: The ratio of the number of pixels displaying a single blue color on the acquisition picture to the number of all pixels on the acquisition picture is less than 20%.

18. The display uniformity compensation method based on machine learning according to claim 13 or 14, characterized in that: The ratio of the number of pixels showing mixed colors on the acquisition picture to the number of all pixels on the acquisition picture is 10% to 30%.

19. A storage medium having computer instructions stored thereon, characterized in that: When the computer instructions are executed, the steps of the method according to any one of claims 1 to 18 are executed.

20. A terminal comprising a memory and a processor, wherein the memory stores computer instructions that can be executed on the processor, characterized in that: When the processor runs the computer instructions, the steps of the method according to any one of claims 1 to 18 are performed.

Images