Screen brightness adjusting method and device, vehicle and equipment

Abstract

The invention provides a screen brightness adjusting method, a screen brightness adjusting device, a vehicle and equipment, which are used for realizing self-adaptive adjustment on screen brightness under different themes. The screen brightness adjusting method comprises the steps of obtaining environment data of an environment where a screen is located currently; adopting a preset regression model to predict target brightness corresponding to the environment data; adaptively adjusting the screen brightness to the target brightness; wherein when the self-learning mode is started, self-learning is carried out according to real-time parameters obtained when the user manually adjusts the screen brightness and the screen feature parameters of the screen under the current theme, and the preset regression model is obtained.

Description

Technical Field

[0001] The present invention relates to the field of screen brightness adjustment for electronic equipment, and in particular to a method, device, vehicle and equipment for adjusting screen brightness. Background Art

[0002] Screen brightness is a key factor in screen display effects and directly affects the user's visual experience. Currently, most central control car computers require manual adjustment of screen brightness, which means that in order to get the best experience, users need to manually adjust the screen brightness under different themes, which is cumbersome. Summary of the invention

[0003] The present invention provides a method, device, vehicle and equipment for adjusting screen brightness, which are used to realize adaptive adjustment of screen brightness under different themes.

[0004] The technical solution of the present invention is:

[0005] On the one hand, the present application provides a method for adjusting the brightness of a screen, comprising:

[0006] Get the environmental data of the current environment of the screen;

[0007] Using a preset regression model, predicting a target brightness corresponding to the environmental data;

[0008] Adaptively adjusting the screen brightness to the target brightness;

[0009] When the self-learning mode is turned on, the preset regression model is obtained by self-learning based on the real-time parameters obtained when the user manually adjusts the screen brightness and the screen feature parameters of the screen under the current theme.

[0010] Preferably, when the self-learning mode is turned on, the step of performing self-learning to obtain the preset regression model according to the real-time parameters obtained when the user manually adjusts the screen brightness and the screen feature parameters of the screen under the current theme includes:

[0011] Determine the expected screen brightness according to the UI value of the brightness bar of the screen manually adjusted by the user, and obtain the real-time environmental data when the user manually adjusts the UI value of the brightness bar of the screen; the real-time parameters include: the expected screen brightness and the real-time environmental data;

[0012] The preset regression model is self-learned and corrected according to the real-time environmental data, the expected screen brightness and the screen characteristic parameters of the screen under the current theme.

[0013] Preferably, the step of performing self-learning correction on the preset regression model according to the real-time environmental data, the expected screen brightness and the screen characteristic parameters of the screen under the current theme includes:

[0014] The real-time environment data and the screen feature parameters of the screen under the current theme are combined to obtain an input feature vector;

[0015] Inputting the input feature vector into the preset regression model to obtain predicted screen brightness;

[0016] Constructing a square sum loss function based on the error between the predicted screen brightness and the expected screen brightness, and calculating the square sum of the error;

[0017] The model coefficients of the preset regression model are corrected according to the sum of squares of the errors.

[0018] Preferably, the screen feature parameters of the screen under the current theme are extracted by a convolutional neural network;

[0019] Each time the screen switches themes, all theme images related to the new theme are extracted from the theme resource library, and all theme images are spliced ​​and processed to obtain a spliced ​​image of a set pixel size;

[0020] Inputting the spliced ​​image into a convolutional neural network to extract screen features;

[0021] The screen features extracted by the convolutional neural network are spliced ​​with the average grayscale and average contrast of the spliced ​​image to obtain the screen feature parameters of the screen under the new theme.

[0022] Preferably, the convolutional neural network comprises: a first convolutional layer, a second convolutional layer, a first pooling layer, a third convolutional layer and a second pooling layer;

[0023] The first and second convolutional layers both use two 5×5 convolution kernels, with a convolution step of 4, and zero padding is performed on the border after each convolution to keep the image size unchanged;

[0024] The first pooling layer uses the maximum pooling with a convolution kernel size of 2×2 and a stride of 2;

[0025] The third convolutional layer uses two 5×5 convolution kernels with a convolution step of 2, and pads the borders with zeros after each convolution to keep the image size unchanged;

[0026] The second pooling layer uses global average pooling.

[0027] Preferably, the step of determining the expected screen brightness according to the UI value of the screen brightness bar manually adjusted by the user comprises:

[0028] Determine the target brightness gradient at which the user manually adjusts the brightness bar UI value of the screen;

[0029] Determine the screen brightness value and the minimum screen brightness UI value corresponding to the target brightness gradient;

[0030] Determining a screen brightness value corresponding to a next brightness gradient after the target brightness gradient;

[0031] Determine the screen brightness increase amount expected by the user under the target brightness gradient according to the screen brightness value corresponding to the target brightness gradient, the minimum screen brightness UI value, and the screen brightness value corresponding to the next brightness gradient;

[0032] The expected screen brightness is determined according to the screen brightness value corresponding to the target brightness gradient and the user's expected screen brightness increase under the target brightness gradient.

[0033] Preferably, before the step of calculating the target brightness corresponding to the environmental data according to a preset regression model, the method further includes:

[0034] The environmental data is filtered using a limited average filtering algorithm.

[0035] On the other hand, the present application also provides a screen brightness adjustment device, comprising:

[0036] The vehicle sensing module is used to obtain the environmental data of the current environment of the screen;

[0037] A regression module, used to predict a target brightness corresponding to the environmental data using a preset regression model;

[0038] An adjustment module, used to adaptively adjust the screen brightness to the target brightness;

[0039] When the self-learning mode is turned on, the preset regression model is obtained by self-learning based on the real-time parameters obtained when the user manually adjusts the screen brightness and the screen feature parameters of the screen under the current theme.

[0040] Preferably, when the self-learning mode is turned on, the step of performing self-learning to obtain the preset regression model according to the real-time parameters obtained when the user manually adjusts the screen brightness and the screen feature parameters of the screen under the current theme includes:

[0041] Determine the expected screen brightness according to the UI value of the brightness bar of the screen manually adjusted by the user, and obtain the real-time environmental data when the user manually adjusts the UI value of the brightness bar of the screen; the real-time parameters include: the expected screen brightness and the real-time environmental data;

[0042] The preset regression model is self-learned and corrected according to the real-time environmental data, the expected screen brightness and the screen characteristic parameters of the screen under the current theme.

[0043] Preferably, the step of performing self-learning correction on the preset regression model according to the real-time environmental data, the expected screen brightness and the screen characteristic parameters of the screen under the current theme includes:

[0044] The real-time environment data and the screen feature parameters of the screen under the current theme are combined to obtain an input feature vector;

[0045] Inputting the input feature vector into a multi-layer perceptron, performing feature extraction and fusion through the hidden layer of the multi-layer perceptron, and obtaining fusion features;

[0046] Inputting the fusion feature into the preset regression model to obtain predicted screen brightness;

[0047] Constructing a square sum loss function based on the error between the predicted screen brightness and the expected screen brightness, and calculating the square sum of the error;

[0048] The model coefficients of the preset regression model are corrected according to the sum of squares of the errors.

[0049] Preferably, the screen feature parameters of the screen under the current theme are extracted by a convolutional neural network;

[0050] Each time the screen switches themes, all theme images related to the new theme are extracted from the theme resource library, and all theme images are spliced ​​and processed to obtain a spliced ​​image of a set pixel size;

[0051] Inputting the spliced ​​image into a convolutional neural network to extract screen features;

[0052] The screen features extracted by the convolutional neural network are spliced ​​with the average grayscale and average contrast of the spliced ​​image to obtain the screen feature parameters of the screen under the new theme.

[0053] Preferably, the convolutional neural network comprises: a first convolutional layer, a second convolutional layer, a first pooling layer, a third convolutional layer and a second pooling layer;

[0054] The first and second convolutional layers both use two 5×5 convolution kernels, with a convolution step of 4, and zero padding is performed on the border after each convolution to keep the image size unchanged;

[0055] The first pooling layer uses the maximum pooling with a convolution kernel size of 2×2 and a stride of 2;

[0056] The third convolutional layer uses two 5×5 convolution kernels with a convolution step of 2, and pads the borders with zeros after each convolution to keep the image size unchanged;

[0057] The second pooling layer uses global average pooling.

[0058] Preferably, the step of determining the expected screen brightness according to the UI value of the screen brightness bar manually adjusted by the user comprises:

[0059] Determine the target brightness gradient at which the user manually adjusts the brightness bar UI value of the screen;

[0060] Determine the screen brightness value and the minimum screen brightness UI value corresponding to the target brightness gradient;

[0061] Determine a screen brightness value corresponding to a next brightness gradient after the target brightness gradient;

[0062] Determine the screen brightness increase amount expected by the user under the target brightness gradient according to the screen brightness value corresponding to the target brightness gradient, the minimum screen brightness UI value, and the screen brightness value corresponding to the next brightness gradient;

[0063] The expected screen brightness is determined according to the screen brightness value corresponding to the target brightness gradient and the user's expected screen brightness increase under the target brightness gradient.

[0064] Preferably, the device further comprises:

[0065] The filtering module is used to filter the environmental data by using a limited average filtering algorithm.

[0066] On the other hand, the present application also provides a vehicle, comprising the above-mentioned screen brightness adjustment device.

[0067] On the other hand, the present application also provides an electronic device, including a processor and a memory, wherein the memory stores a computer program, and the processor executes the screen brightness adjustment method as described above by calling the computer program.

[0068] The beneficial effects of the present invention are:

[0069] Since the current theme of the screen and the environmental data of the environment in which the screen is currently located (such as the ambient light brightness) will affect the user's subjective perception of the screen brightness; when the user manually adjusts the screen brightness, it is inevitable to adjust the screen brightness in combination with the current theme of the screen and the current environmental parameters (such as the ambient light brightness). Therefore, in this application, the preferences of different users for screen brightness under different themes and different environmental parameters are taken into consideration, and the real-time parameters obtained when the user manually adjusts the screen brightness and the screen characteristic parameters of the screen under the current theme are combined for self-learning to obtain a preset regression model, and the target brightness predicted by the preset regression model is more in line with the user's preference. BRIEF DESCRIPTION OF THE DRAWINGS

[0070] Figure 1 Schematic diagram of a method for adjusting screen brightness in an embodiment of the present invention Figure 1;

[0071] Figure 2 Schematic diagram of a method for adjusting screen brightness in an embodiment of the present invention Figure 2 ;

[0072] Figure 3 Schematic diagram of a method for adjusting screen brightness in an embodiment of the present invention Figure 3 ;

[0073] Figure 4 Schematic diagram of a screen brightness adjustment device in an embodiment of the present invention. DETAILED DESCRIPTION

[0074] In order to make the purpose, technical scheme and advantages of the present invention clearer, the present invention is further described in detail below in conjunction with specific embodiments and with reference to the accompanying drawings. It should be understood that these descriptions are only exemplary and are not intended to limit the scope of the present invention. In addition, in the following description, the description of well-known structures and technologies is omitted to avoid unnecessary confusion of the concept of the present invention.

[0075] like Figure 1 In this embodiment, a method for adjusting the brightness of a screen is provided, comprising:

[0076] S1, obtaining the environmental data of the current environment of the screen;

[0077] S2, using the preset regression model to predict the target brightness corresponding to the environmental data;

[0078] S3, adaptively adjust the screen brightness to the target brightness.

[0079] In the embodiment of the present application, the screen may be a screen of an electronic device, such as a mobile phone, tablet, computer, television, etc., or it may be a screen of a vehicle computer mounted on a vehicle.

[0080] The environmental data of the current environment of the screen may include at least the ambient light brightness, and may also include the ambient light brightness and humidity. These data may be collected by an ambient light brightness sensor and a humidity sensor mounted on an electronic device, or by using an ambient light brightness sensor and a humidity sensor mounted on a vehicle.

[0081] The preset regression model in the embodiment of the present application is obtained by self-learning based on the historical brightness parameters of the screen brightness manually adjusted by the user and the screen feature parameters when the screen last switched themes. The input of the preset regression model is the fusion feature obtained by feature extraction and fusion based on the historical brightness parameters of the screen brightness manually adjusted by the user and the screen feature parameters when the screen last switched themes, and the output of the preset regression model is the target brightness of the screen.

[0082] The regression algorithm used by the preset regression model includes, but is not limited to, linear regression, logistic regression, polynomial regression, stepwise regression, ridge regression, Lasso regression, etc. In this embodiment, the preset regression model is taken as a regression tree model as an example to explain the initial model and the self-learning process of the preset regression model. First, the initial values ​​of the coefficients of the regression tree model need to be determined before the initial model of the regression tree model can be obtained.

[0083] In an embodiment of the present application, environmental data or conclusions about human perception, reaction, comfort, etc. output by human factors engineering are used as part of the input data for constructing an initial model of a regression tree model. That is, human factors engineering is used to calibrate the screen brightness that makes people feel comfortable under different themes and different environmental data. For example, when the environmental data is ambient light brightness, under the daytime theme, the obtained ambient light brightness-expected screen brightness mapping table is referred to in Table 1 below.

[0084]

[0085] Table 1

[0086] In addition, theme skin images related to different themes are stored in the theme resource library of the device, and all theme skin images related to the theme selected in the aforementioned human factors engineering are read from the theme resource library. For example, if the theme in Table 1 includes a daytime theme, then theme skin images related to the daytime theme are selected from the theme resource library of the device.

[0087] Since there are multiple skin images related to different themes, in order to facilitate data processing, it is necessary to splice all skin images related to each theme, and make the obtained spliced ​​image have a fixed pixel size through processing methods such as color filling or scaling.

[0088] Furthermore, the spliced ​​images with a fixed pixel size are subjected to feature extraction through a convolutional neural network to obtain the screen feature parameters under the daytime theme.

[0089] The convolutional neural network used in this embodiment includes: a first convolutional layer, a second convolutional layer, a first pooling layer, a third convolutional layer and a second pooling layer;

[0090] The first and second convolutional layers both use two 5×5 convolution kernels, with a convolution step of 4, and zero padding is performed on the border after each convolution to keep the image size unchanged;

[0091] The first pooling layer uses the maximum pooling with a convolution kernel size of 2×2 and a stride of 2;

[0092] The third convolutional layer uses two 5×5 convolution kernels with a convolution step of 2, and pads the borders with zeros after each convolution to keep the image size unchanged;

[0093] The second pooling layer uses global average pooling.

[0094] Therefore, assuming that the pixel size of the stitched image is 1920*1080, the dimension of the tensor input to the convolutional neural network is (1920,1080,1,3).

[0095] Considering that grayscale and contrast have a great influence on screen brightness, in the embodiment of the present application, two 5x5 convolution kernel initialization parameters are designed for the first convolution layer, the second convolution layer, and the third convolution layer, one of which is a mean filter K1, which is used for smoothing, denoising, and extracting overall grayscale features:

[0096]

[0097] One is a sharpening filter K2, which is used to extract edge and texture features to reflect image contrast:

[0098]

[0099] Therefore, after being processed by the second pooling layer of the convolutional neural network, the tensor (1,6) is obtained.

[0100] Grayscale and contrast are also very important features, and these simple features can be used to reduce the model complexity of the convolutional neural network, smooth the model, and prevent overfitting. The average grayscale and contrast calculation results of the spliced ​​image are spliced ​​into a tensor (1,2), and the tensor (1,6) obtained after the second pooling layer is spliced ​​to obtain a tensor of dimension (1,8) as the screen feature parameters of the final output of the convolutional neural network; among them, the 4-dimensional final output of the convolutional neural network represents the grayscale, contrast, and features extracted by the two convolution kernels of each image.

[0101] Furthermore, the brightness of each ambient light under the daytime theme is collected as a tensor of (1,1), which is concatenated with the tensor of dimension (1,8) finally output by the convolutional neural network to obtain an input tensor of (1,9), which is used as the input feature vector.

[0102] The input feature vector (1,9) is input into the pre-established regression tree model, that is, the input feature vector (1,9) is used as the input parameter X in the regression tree model, and the expected screen brightness is used as the output parameter According to multiple groups (X, ) can get the initial values ​​of each coefficient in the regression tree model, and the initial model of the regression tree model can be obtained.

[0103] After the initial model of the regression tree model is constructed, the initial model can be used to predict the screen brightness.

[0104] As users use electronic devices for longer periods of time, their usage habits may change. Under the same theme, users may also adjust the screen brightness. Therefore, after a period of use, the historical brightness parameters collected when users manually adjust the screen brightness and the screen feature parameters under the current theme can be used as training data to retrain the initial model of the preset regression model, thereby updating the model coefficients. Figure 2 At this time, the screen brightness adjustment method in this embodiment further includes:

[0105] S4, when the self-learning mode is turned on, self-learning is performed to obtain a preset regression model based on the real-time parameters obtained when the user manually adjusts the screen brightness and the screen feature parameters of the screen under the current theme.

[0106] In the embodiment of the present application, when the self-learning mode is turned on, the step S4 of self-learning to obtain a preset regression model according to the real-time parameters obtained when the user manually adjusts the screen brightness and the screen feature parameters of the screen under the current theme includes:

[0107] Determine the expected screen brightness according to the UI value of the brightness bar of the screen manually adjusted by the user, and obtain the real-time environmental data when the user manually adjusts the UI value of the brightness bar of the screen; the real-time parameters include: the expected screen brightness and the real-time environmental data;

[0108] The preset regression model is self-learned and corrected according to the real-time environmental data, the expected screen brightness and the screen characteristic parameters of the screen under the current theme.

[0109] In this embodiment, the step of determining the expected screen brightness according to the UI value of the screen brightness bar manually adjusted by the user includes:

[0110] When the user manually adjusts the screen brightness bar UI value to 0, determining that the expected screen brightness is a first preset value;

[0111] When the user manually adjusts the screen brightness bar UI value to be greater than 100, determining that the expected screen brightness is a second preset value;

[0112] When the UI value of the brightness bar of the screen manually adjusted by the user is not 0 or greater than 100, determining the target brightness gradient of the UI value of the brightness bar of the screen manually adjusted by the user;

[0113] Determine the screen brightness value and the minimum screen brightness UI value corresponding to the target brightness gradient;

[0114] Determine a screen brightness value corresponding to a next brightness gradient after the target brightness gradient;

[0115] Determine the expected screen brightness increase amount according to the screen brightness value corresponding to the target brightness gradient and the minimum screen brightness UI value and the screen brightness value corresponding to the next brightness gradient;

[0116] The expected screen brightness is determined according to the screen brightness value corresponding to the target brightness gradient and the expected screen brightness increase.

[0117] Furthermore, the step of performing self-learning correction on the preset regression model according to the real-time environmental data, the expected screen brightness and the screen characteristic parameters of the screen under the current theme includes:

[0118] The real-time environment data and the screen feature parameters of the screen under the current theme are combined to obtain an input feature vector;

[0119] The concatenated input feature vector is input into the preset regression model to obtain the predicted screen brightness;

[0120] Constructing a square sum loss function based on the error between the predicted screen brightness and the expected screen brightness, and calculating the square sum of the error;

[0121] Then, the model coefficients of the preset regression model are corrected according to the obtained sum of squares of the errors.

[0122] Still taking the preset regression model as a regression tree model as an example, when updating the preset regression model, the overall logic is consistent with the aforementioned initial model construction logic. The only difference is that at this time, the expected screen brightness is determined based on the UI value of the screen brightness bar manually adjusted by the user, rather than the expected screen brightness based on human factors engineering.

[0123] Among them, the aforementioned brightness bar UI value specifically refers to the brightness bar UI value intuitively displayed on the screen. The brightness bar UI value currently generally adopts a range value of 0 to 100, and the brightness bar UI value of 0 to 100 is divided into multiple brightness gradients. Different brightness gradients are mapped to the screen brightness. The user adjusts the brightness bar UI value intuitively, which is actually adjusting the brightness gradient and the corresponding screen brightness. The brightness bar UI value is set to facilitate intuitive operation by the operator.

[0124] In the embodiment of the present application, the brightness bar UI value of 0 to 100 is divided into multiple brightness gradients. Figure 1, set the brightness gradient to 11. Among them, the first brightness gradient contains a brightness bar UI value of 0; the second brightness gradient contains a brightness bar UI value of 10, corresponding to 1-10; the third brightness gradient contains a brightness bar UI value of 10, corresponding to 11-20; the fourth brightness gradient contains a brightness bar UI value of 10, corresponding to 21-30; the fifth brightness gradient contains a brightness bar UI value of 10, corresponding to 31-40; the sixth brightness gradient contains a brightness bar UI value of 10, corresponding to 41-50; the seventh brightness gradient contains a brightness bar UI value of 10, corresponding to 51-60; the eighth brightness gradient contains a brightness bar UI value of 10, corresponding to 61-70; the ninth brightness gradient contains a brightness bar UI value of 10, corresponding to 71-80; the tenth brightness gradient contains a brightness bar UI value of 10, corresponding to 81-90; the eleventh brightness gradient contains a brightness bar UI value of 10, corresponding to 91-100.

[0125] The specific implementation method of determining the expected screen brightness when the user manually adjusts the screen brightness bar UI value to not 0 or greater than 100 is described as follows: Assuming that the user manually adjusts the screen brightness bar UI value to 55%, the target brightness gradient is determined to be the sixth brightness gradient; thus, the minimum screen brightness UI value of the sixth brightness gradient is determined to be 50%, and the screen brightness value corresponding to the sixth brightness gradient is 290cd / m 2 Furthermore, when the target brightness gradient is the sixth brightness gradient, the next brightness gradient is the seventh brightness gradient. Combined with Table 1, the screen brightness value corresponding to the seventh brightness gradient is 430 cd / m 2 .

[0126] Furthermore, the difference between the screen brightness value corresponding to the seventh brightness gradient and the screen brightness value corresponding to the sixth brightness gradient is calculated, and the difference between the brightness bar UI value of the screen manually adjusted by the user and the minimum screen brightness UI value corresponding to the sixth brightness gradient is calculated, and then the user's expected screen brightness increase under the sixth brightness gradient is determined based on the screen brightness difference and the UI value difference, and then the expected screen brightness is determined based on the screen brightness value corresponding to the sixth brightness gradient and the expected screen brightness increase.

[0127] In the actual process, the expected screen brightness S can be obtained by simply inputting the UI value u of the screen brightness bar manually adjusted by the user into the following preset formula; the preset formula is:

[0128]

[0129] l0 is the first preset value;

[0130] l n is a second preset value;

[0131] l i The screen brightness value corresponding to the i-th brightness gradient where the UI value u of the screen brightness bar manually adjusted by the user is located;

[0132] l i+1 Take the screen brightness value corresponding to the i+1th brightness gradient;

[0133] p represents the number of UI values ​​contained between two adjacent brightness gradients of the screen, and p is a preset value;

[0134] ui represents the minimum screen brightness bar UI value corresponding to the i-th brightness gradient of the screen;

[0135] u(i+1) represents the minimum screen brightness bar UI value corresponding to the i+1th brightness gradient of the screen.

[0136] Taking the preset regression model as a regression tree model as an example, the process of modifying the regression tree model under the daytime theme includes the following steps:

[0137] When Y = 0, initialize the current coefficient w of the regression tree model i The value is 0;

[0138] When Y≠0, initialize the coefficient w of the regression tree model i is the initial value obtained above;

[0139] Then, using the current coefficient value w i , the corresponding ambient light brightness when the user manually adjusts the screen brightness bar UI value is input into the regression tree model to obtain the predicted screen brightness

[0140] Then, calculate each predicted screen brightness The error between the expected screen brightness y and

[0141] The sum of squares of the errors is calculated using the sum of squares loss function e i is the error calculated based on the UI value of the ith brightness bar manually adjusted by the user; n is the total number of UI values ​​of the brightness bar manually adjusted by the user;

[0142] Then, the coefficients are updated according to the square sum of the errors, where, represents the updated coefficient, w i is the current coefficient value, α is the learning rate of the model, which is a preset fixed value;

[0143] Thus, a self-learning corrected regression tree model is obtained.

[0144] Taking into account the fluctuations and abnormal values ​​of the light sensor, the environmental data of the environment in which the screen is located is also preprocessed in this embodiment. Among them, when the environmental data is obtained, due to the instability of the sensor data, the screen brightness is directly determined according to the environmental data collected by the sensor, and there is a possibility of screen brightness jump. For this reason, in this embodiment, when the environmental data is obtained, a limited average filtering algorithm is used to filter it. Of course, other filtering methods can also be used for filtering, such as median filtering, arithmetic mean filtering, recursive average filtering, median average filtering, first-order lag filtering and other filtering methods. Specifically, when the limited average filtering algorithm is used to filter the ambient light brightness, each time a new ambient light brightness is obtained, the environmental data values ​​reported by the sensor for the previous N times (N is set to 5 in this embodiment) are selected, and the standard deviation of the environmental data values ​​for the previous N times is set to σ, and the maximum amplitude ζ is set to 3σ. Then through the formula:

[0145]

[0146] The output is the ambient light brightness b[n] that is finally output to the preset regression model, where

[0147]

[0148] a[n]: the nth input value

[0149] b[n]: nth output value

[0150] t: period or sliding window length. In this embodiment, t=5

[0151] The average value of the input value within the period

[0152] T: filtering algorithm threshold;

[0153] Represents the standard deviation σ of the environmental data values ​​of the previous 5 times.

[0154] The screen brightness bar UI value mentioned in the above description is not only for users to actively adjust, but also adaptively adjusted according to the target brightness when the screen brightness is adaptively adjusted. Among them, the screen brightness s and the brightness bar UI value U have the following corresponding relationship:

[0155]

[0156] When the screen brightness S changes, the real-time value of the brightness bar UI value can be determined according to the above formula, so as to adaptively adjust the brightness bar UI value.

[0157] Reference Figure 3 ,Should Figure 3Detailed definition of how to correct the preset regression model in the night mode after the user manually adjusts the UI value of the screen brightness bar and the application scenarios of adjusting the screen brightness in the night mode based on the corrected preset regression model. Figure 3 In the process, the following steps are performed:

[0158] S101, determine whether the user manually adjusts the UI value of the brightness bar of the screen. If adjusted, proceed to step S102.

[0159] S102, using the formula in the above description, determine the expected screen brightness according to the UI value of the brightness bar manually adjusted by the user.

[0160] S103, adaptively adjusting the screen brightness to the desired screen brightness.

[0161] The aforementioned steps S101 to S103 enable the user to manually adjust the screen brightness.

[0162] After the user manually adjusts the screen brightness, the following operations are also performed:

[0163] S104, determine whether the preset regression self-learning mode is turned on. If it is turned on, enter S105.

[0164] Among them, the preset regression self-learning mode can be actively enabled by the user, or automatically enabled when the user manually adjusts the screen brightness a preset number of times.

[0165] S105, performing self-learning correction on the preset regression model according to the real-time environmental data corresponding to the UI value of the brightness bar of the screen manually adjusted by the user, the expected screen brightness, and the screen characteristic parameters of the screen under the current theme.

[0166] Based on steps S104 and S105, self-learning correction of the preset regression model is implemented, wherein the self-learning correction process refers to the description of the aforementioned content.

[0167] Then, the preset regression model after self-learning correction can be used to adjust the screen brightness, that is, execute:

[0168] S106, obtaining environmental data and setting the display theme of the screen to a night theme.

[0169] S107, calculating the target mode corresponding to the ambient light brightness according to the modified preset regression model.

[0170] S108, adaptively adjusting the screen brightness to the target brightness.

[0171] S109, adjusting the screen brightness bar UI value according to the target brightness.

[0172] Therefore, according to S106 to S109, the screen brightness in the night mode is adaptively adjusted using the preset regression model corrected by self-learning.

[0173] The above screen brightness adjustment method can be applied to all electronic screens, such as mobile phones, tablets, computers, vehicles, medical equipment, aviation equipment and other equipment.

[0174] Since the current theme of the screen and the environmental data of the environment in which the screen is currently located (such as the ambient light brightness) will affect the user's subjective perception of the screen brightness; when the user manually adjusts the screen brightness, it is inevitable to adjust the screen brightness in combination with the current theme of the screen and the current environmental parameters (such as the ambient light brightness). Therefore, in this application, the preferences of different users for screen brightness under different themes and different environmental parameters are taken into consideration, and the real-time parameters obtained when the user manually adjusts the screen brightness and the screen characteristic parameters of the screen under the current theme are combined for self-learning to obtain a preset regression model, and the target brightness predicted by the preset regression model is more in line with the user's preference.

[0175] On the other hand, the present application also provides a screen brightness adjustment device, referring to Figure 4 , the device comprises:

[0176] The vehicle sensing module 101 is used to obtain the environmental data of the current environment of the screen;

[0177] A regression module 102 is used to predict a target brightness corresponding to the environmental data using a preset regression model;

[0178] An adjustment module 103, used to adaptively adjust the screen brightness to the target brightness;

[0179] When the self-learning mode is turned on, the preset regression model is obtained by self-learning based on the real-time parameters obtained when the user manually adjusts the screen brightness and the screen feature parameters of the screen under the current theme.

[0180] Preferably, when the self-learning mode is turned on, the step of performing self-learning to obtain the preset regression model according to the real-time parameters obtained when the user manually adjusts the screen brightness and the screen feature parameters of the screen under the current theme includes:

[0181] Determine the expected screen brightness according to the UI value of the brightness bar of the screen manually adjusted by the user, and obtain the real-time environmental data when the user manually adjusts the UI value of the brightness bar of the screen; the real-time parameters include: the expected screen brightness and the real-time environmental data;

[0182] The preset regression model is self-learned and corrected according to the real-time environmental data, the expected screen brightness and the screen characteristic parameters of the screen under the current theme.

[0183] Preferably, the step of performing self-learning correction on the preset regression model according to the real-time environmental data, the expected screen brightness and the screen characteristic parameters of the screen under the current theme includes:

[0184] The real-time environment data and the screen feature parameters of the screen under the current theme are combined to obtain an input feature vector;

[0185] The input feature vector is input into a preset regression model to obtain a predicted screen brightness;

[0186] Constructing a square sum loss function based on the error between the predicted screen brightness and the expected screen brightness, and calculating the square sum of the error;

[0187] The model coefficients of the preset regression model are corrected according to the sum of squares of the errors.

[0188] Preferably, the screen feature parameters of the screen under the current theme are extracted by a convolutional neural network;

[0189] Each time the screen switches themes, all theme images related to the new theme are extracted from the theme resource library, and all theme images are spliced ​​and processed to obtain a spliced ​​image of a set pixel size;

[0190] Inputting the spliced ​​image into a convolutional neural network to extract screen features;

[0191] The screen features extracted by the convolutional neural network are spliced ​​with the average grayscale and average contrast of the spliced ​​image to obtain the screen feature parameters of the screen under the new theme.

[0192] Preferably, the convolutional neural network comprises: a first convolutional layer, a second convolutional layer, a first pooling layer, a third convolutional layer and a second pooling layer;

[0193] The first and second convolutional layers both use two 5×5 convolution kernels, with a convolution step of 4, and zero padding is performed on the border after each convolution to keep the image size unchanged;

[0194] The first pooling layer uses the maximum pooling with a convolution kernel size of 2×2 and a stride of 2;

[0195] The third convolutional layer uses two 5×5 convolution kernels with a convolution step of 2, and pads the borders with zeros after each convolution to keep the image size unchanged;

[0196] The second pooling layer uses global average pooling.

[0197] Preferably, the step of determining the expected screen brightness according to the UI value of the screen brightness bar manually adjusted by the user comprises:

[0198] Determine the target brightness gradient at which the user manually adjusts the brightness bar UI value of the screen;

[0199] Determine the screen brightness value and the minimum screen brightness UI value corresponding to the target brightness gradient;

[0200] Determine a screen brightness value corresponding to a next brightness gradient after the target brightness gradient;

[0201] Determine the screen brightness increase amount expected by the user under the target brightness gradient according to the screen brightness value corresponding to the target brightness gradient, the minimum screen brightness UI value, and the screen brightness value corresponding to the next brightness gradient;

[0202] The expected screen brightness is determined according to the screen brightness value corresponding to the target brightness gradient and the user's expected screen brightness increase under the target brightness gradient.

[0203] Preferably, before the step of calculating the target brightness corresponding to the environmental data according to a preset regression model, the device further includes:

[0204] The filtering module uses a limited average filtering algorithm to filter the environmental data.

[0205] Since the current theme of the screen and the environmental data of the environment in which the screen is currently located (such as the ambient light brightness) will affect the user's subjective perception of the screen brightness; when the user manually adjusts the screen brightness, it is inevitable to adjust the screen brightness in combination with the current theme of the screen and the current environmental parameters (such as the ambient light brightness). Therefore, in this application, the preferences of different users for screen brightness under different themes and different environmental parameters are taken into consideration, and the real-time parameters obtained when the user manually adjusts the screen brightness and the screen characteristic parameters of the screen under the current theme are combined for self-learning to obtain a preset regression model, and the target brightness predicted by the preset regression model is more in line with the user's preference.

[0206] An embodiment of the present invention further provides a vehicle, comprising the above-mentioned screen brightness adjustment device.

[0207] An embodiment of the present invention further provides a storage medium on which a computer program is stored. When the computer program is executed on a computer, the computer executes the screen brightness adjustment method as described above.

[0208] An embodiment of the present invention further provides an electronic device, including a processor and a memory, wherein the memory stores a computer program, and the processor is used to execute the above-mentioned screen brightness adjustment method by calling the computer program.

[0209] It should be noted that the various embodiments in this specification are described in a progressive manner, and each embodiment focuses on the differences from other embodiments. The same or similar parts between the various embodiments can be referenced to each other.

[0210] Although the preferred embodiments of the present invention have been described, those skilled in the art may make additional changes and modifications to these embodiments once they have learned the basic creative concept. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments and all changes and modifications that fall within the scope of the embodiments of the present invention.

[0211] It should also be noted that, in this article, the orientation or position relationship indicated by the terms "center", "up", "down", "left", "right", "vertical", "horizontal", "inside", "outside", etc. is based on the orientation or position relationship shown in the drawings, which is for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore cannot be understood as a limitation of the present invention. In addition, relational terms such as "first" and "second" are used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply that there is any such actual relationship or order between these entities or operations, nor can they be understood as indicating or implying relative importance. Moreover, the terms "include", "comprise" or any other variant thereof are intended to cover non-exclusive inclusion, so that a process, method, article or terminal device including a series of elements does not include those elements, but also includes other elements not explicitly listed, or also includes elements inherent to such a process, method, article or terminal device. In the absence of further restrictions, the elements defined by the sentence "including one..." do not exclude the existence of other identical elements in the process, method, article or terminal device including the elements.

[0212] The technical solution provided by the present invention is described in detail above. The principle and implementation mode of the present invention are described in this article using specific examples. The description of the above embodiments is only used to help understand the present invention, and the content of this specification should not be understood as limiting the present invention. At the same time, for those skilled in the art, according to the present invention, there will be different forms of changes in the specific implementation mode and application scope. It is not necessary and impossible to list all the implementation modes here, and the obvious changes or modifications derived therefrom are still within the protection scope of the present invention.

Claims

1. A method for adjusting screen brightness, characterized in that: include: Get the environmental data of the current environment of the screen; Using a preset regression model, predicting a target brightness corresponding to the environmental data; Adaptively adjusting the screen brightness to the target brightness; When the self-learning mode is turned on, the preset regression model is obtained by self-learning based on the real-time parameters obtained when the user manually adjusts the screen brightness and the screen feature parameters of the screen under the current theme.

2. The screen brightness adjustment method according to claim 1, characterized in that: When the self-learning mode is turned on, the step of performing self-learning to obtain the preset regression model according to the real-time parameters obtained when the user manually adjusts the screen brightness and the screen feature parameters of the screen under the current theme includes: Determine the expected screen brightness according to the UI value of the brightness bar of the screen manually adjusted by the user, and obtain the real-time environmental data when the user manually adjusts the UI value of the brightness bar of the screen; the real-time parameters include: the expected screen brightness and the real-time environmental data; The preset regression model is self-learned and corrected according to the real-time environmental data, the expected screen brightness and the screen characteristic parameters of the screen under the current theme.

3. The screen brightness adjustment method according to claim 2, characterized in that: The step of performing self-learning correction on the preset regression model according to the real-time environmental data, the expected screen brightness and the screen characteristic parameters of the screen under the current theme includes: The real-time environment data and the screen feature parameters of the screen under the current theme are combined to obtain an input feature vector; Inputting the input feature vector into the preset regression model to obtain predicted screen brightness; Constructing a square sum loss function based on the error between the predicted screen brightness and the expected screen brightness, and calculating the square sum of the error; The model coefficients of the preset regression model are corrected according to the sum of squares of the errors.

4. The method for adjusting screen brightness according to claim 2, characterized in that: The screen feature parameters of the screen under the current theme are extracted through convolutional neural network; Each time the screen switches themes, all theme images related to the new theme are extracted from the theme resource library, and all theme images are spliced ​​and processed to obtain a spliced ​​image of a set pixel size; Inputting the spliced ​​image into a convolutional neural network to extract screen features; The screen features extracted by the convolutional neural network are spliced ​​with the average grayscale and average contrast of the spliced ​​image to obtain the screen feature parameters of the screen under the new theme.

5. The screen brightness adjustment method according to claim 4, characterized in that: The convolutional neural network comprises: a first convolutional layer, a second convolutional layer, a first pooling layer, a third convolutional layer and a second pooling layer; The first and second convolutional layers both use two 5×5 convolution kernels, with a convolution step of 4, and zero padding is performed on the border after each convolution to keep the image size unchanged; The first pooling layer uses the maximum pooling with a convolution kernel size of 2×2 and a stride of 2; The third convolutional layer uses two 5×5 convolution kernels with a convolution step of 2, and pads the borders with zeros after each convolution to keep the image size unchanged; The second pooling layer uses global average pooling.

6. The screen brightness adjustment method according to claim 2, characterized in that: The step of determining the expected screen brightness according to the UI value of the screen brightness bar manually adjusted by the user comprises: Determine the target brightness gradient at which the user manually adjusts the brightness bar UI value of the screen; Determine the screen brightness value and the minimum screen brightness UI value corresponding to the target brightness gradient; Determining a screen brightness value corresponding to a next brightness gradient after the target brightness gradient; Determine the screen brightness increase amount expected by the user under the target brightness gradient according to the screen brightness value corresponding to the target brightness gradient, the minimum screen brightness UI value, and the screen brightness value corresponding to the next brightness gradient; The expected screen brightness is determined according to the screen brightness value corresponding to the target brightness gradient and the user's expected screen brightness increase under the target brightness gradient.

7. The screen brightness adjustment method according to claim 1, characterized in that: Before the step of calculating the target brightness corresponding to the environmental data according to the preset regression model, the method further includes: The environmental data is filtered using a limited average filtering algorithm.

8. A screen brightness adjustment device, characterized in that: include: The vehicle sensing module is used to obtain the environmental data of the current environment of the screen; A regression module, used to predict a target brightness corresponding to the environmental data using a preset regression model; An adjustment module, used to adaptively adjust the screen brightness to the target brightness; When the self-learning mode is turned on, the preset regression model is obtained by self-learning based on the real-time parameters obtained when the user manually adjusts the screen brightness and the screen feature parameters of the screen under the current theme.

9. A vehicle, characterized in that: A brightness adjustment device for a screen comprising the device described in claim 8.

10. An electronic device comprising a processor and a memory, wherein the memory stores a computer program, characterized in that: The processor is used to execute the screen brightness adjustment method as described in any one of claims 1 to 7 by calling the computer program.

Images