Mobile phone liquid crystal screen dynamic color temperature adjusting method and system based on AI

Abstract

The invention discloses a mobile phone liquid crystal screen dynamic color temperature adjusting method and system based on AI, and relates to the technical field of mobile phone liquid crystal screen AI color temperature adjusting. Firstly, environment light intensity and image data are collected through an environment light sensor and a camera, image features are extracted, and user preference data are collected; comprising the steps of manually adjusting historical, geographic positions, time information and the like, training a deep learning model by using the data, inputting real-time environment and user preference data into the model to calculate an appropriate color temperature value, sending the appropriate color temperature value to a driving chip to adjust the luminous intensity proportion of three primary colors of a liquid crystal screen, finally collecting a user feedback optimization model, and covering exception handling and energy-saving adaptation. The color temperature of the liquid crystal screen of the mobile phone can be intelligently adjusted, different environments can be accurately adapted, personalized preferences of users are met, comfortable visual experience is brought, the functions of exception handling and energy saving are achieved, use is guaranteed, energy consumption is reduced, and the overall use performance of the mobile phone and the user satisfaction degree are improved.

Description

Technical Field

[0001] The present invention relates to the technical field of AI color temperature adjustment for mobile phone liquid crystal screens, and in particular to an AI-based dynamic color temperature adjustment method and system for mobile phone liquid crystal screens. Background Art

[0002] With the widespread use of smartphones in daily life, users are increasingly demanding the display quality of their phone's LCD screens. Currently, color temperature adjustment for mobile phone LCD screens primarily relies on fixed modes or simple manual adjustments. In fixed mode, mobile phone LCD screens typically offer only a few preset color temperature options, such as cool, warm, and standard. This approach fails to dynamically adjust to the user's actual environment and personal preferences. Under varying lighting conditions, such as bright outdoor environments or dim indoor environments, the screen display quality may be poor, with the image appearing too bright or too dark, or with color distortion, impacting the user's visual experience and comfort.

[0003] While manual adjustment allows users to adjust color temperature based on their preferences, it's a cumbersome process, requiring frequent access to the settings menu. Furthermore, it's difficult for users to accurately determine the most suitable color temperature for a specific environment based on experience, resulting in suboptimal adjustment results. Furthermore, color temperature preferences vary among users, and existing adjustment methods can't effectively record and adapt to these individual preferences, hindering the ability to provide customized display services.

[0004] In recent years, some automatic color temperature adjustment technologies based on ambient light intensity have emerged. However, these technologies only consider the single factor of light intensity, ignoring the impact of environmental scenes, object colors, and other factors on human visual perception. Furthermore, they lack the ability to learn and adapt to users' personalized preferences, making them unable to meet their diverse needs. In actual use, when users quickly switch from one environment to another, the screen color temperature cannot be adjusted promptly and accurately, resulting in a significant lag in the display effect, seriously affecting the user experience. Therefore, there is an urgent need for a dynamic color temperature adjustment method and system for mobile phone LCD screens that can comprehensively consider multiple factors and automatically adapt to user preferences. Summary of the Invention

[0005] The present invention proposes an AI-based dynamic color temperature adjustment method and system for mobile phone LCD screens to solve the problems mentioned in the above-mentioned prior art.

[0006] In order to achieve the above object, the present invention adopts the following technical solutions:

[0007] An AI-based method for dynamic color temperature adjustment of a mobile phone LCD screen includes the following steps:

[0008] Environmental data collection steps: The ambient light intensity data is collected through the mobile phone's ambient light sensor, and the camera collects environmental image data. The ambient light sensor uses a silicon photodiode sensor, and the camera supports autofocus and white balance functions;

[0009] Image feature extraction step: Input the collected environmental image data into the trained convolutional neural network model, extract image feature information through the model, and use the Softmax function Classify and predict object type, x is the input image data, f j is the feature vector, C is the total number of categories, y i is the i-th category;

[0010] User preference data collection steps: Set the user preference interface through the mobile phone application, collect user history data, user stay time data, and user color temperature rating data, and at the same time, obtain the geographic location and time information in the mobile phone system;

[0011] Color temperature adjustment model training steps: ambient light intensity data, image feature data, and user preference data are used as input, and the color temperature value that the user is satisfied with is used as output. The back propagation algorithm and stochastic gradient descent method are used to train the deep learning model, and the loss function uses the mean square error formula. Adjust the model parameters, n is the number of samples, y i is the true color temperature value, To predict the color temperature value;

[0012] Real-time color temperature calculation steps: Input the ambient light intensity data, extracted image feature data, and user preference data into the trained color temperature adjustment model. The LCD screen color temperature value is calculated using the formula T = α × I + β × F + γ × P, where T is the predicted color temperature value, I is the ambient light intensity data, F is the image feature data vector, P is the user preference data vector, and α, β, and γ are weight coefficients.

[0013] Color temperature adjustment execution steps: send the color temperature value to the driver chip of the mobile phone LCD screen, and the driver chip adjusts the luminous intensity ratio of the three primary colors in the LCD screen backlight;

[0014] Feedback optimization step: Collect user feedback data, use the feedback data and original input data as new training samples, and retrain the color temperature adjustment model.

[0015] Furthermore, it also includes an abnormal situation handling step: when the light intensity data collected by the ambient light sensor fluctuates abnormally or the image data collected by the camera cannot be parsed normally, the system automatically switches to the default color temperature adjustment mode, the default color temperature value is 6500K, and a prompt message is displayed on the mobile phone interface.

[0016] Furthermore, the energy-saving mode adaptation step is as follows: when the battery level of the mobile phone is less than 20%, the system automatically reduces the brightness of the LCD screen and increases the threshold range of color temperature adjustment while ensuring the user's visual comfort.

[0017] Furthermore, in the image feature extraction step, the collected environmental image is preprocessed, including grayscale processing, noise reduction processing and normalization processing. The grayscale processing adopts the weighted average method, and the formula is Gray = 0.299×R+0.587×G+0.114×B, where R, G, and B are the red, green, and blue channel values ​​of the image, respectively. The noise reduction processing adopts the median filtering algorithm with a filter window size of 3×3; the normalization processing normalizes the image pixel values ​​to the range of [0,1].

[0018] Furthermore, in the user preference data collection step, the collaborative filtering algorithm is used to analyze the similarity between users, and the cosine similarity formula between users is calculated. Recommend color temperature settings with similar user preferences to users, where I uv is the set of items that users u and v have jointly evaluated, r ui and r vi are the ratings of user u and user v on item i respectively.

[0019] Furthermore, in the color temperature adjustment model training step, the transfer learning method is adopted. The CNN model pre-trained on a large-scale public image dataset is used as a basis. On this basis, it is fine-tuned for the mobile phone LCD screen color temperature adjustment task, reducing training time and data requirements.

[0020] Furthermore, in the feedback optimization step, a reinforcement learning algorithm is introduced, and the user's feedback is used as a reward signal. The long-term cumulative reward is maximized by adjusting the model parameters. The Q-learning algorithm is adopted, and its core formula is Q(s, a)←Q(s, a)+β[r+γmax a′ Q(s′, a′)-Q(sa)], where s is the current state, a is the current action, r is the immediate reward, γ is the discount factor, α is the learning rate, and s' is the next state, continuously optimizing the color temperature adjustment strategy.

[0021] Furthermore, the system for the method of dynamic color temperature adjustment of mobile phone LCD screen based on AI includes the following modules:

[0022] Environmental data acquisition module: It consists of an ambient light sensor and a camera. The ambient light sensor is used to collect ambient light intensity data in real time; the camera is used to collect environmental image data. The module also includes a data cache unit to cache the data collected in the last 10 seconds to prevent data loss.

[0023] Image feature extraction module: It includes an image preprocessing submodule and a feature extraction submodule. The image preprocessing submodule is used to grayscale, reduce noise, and normalize the collected environmental images. The feature extraction submodule uses a convolutional neural network model to extract feature information from the image.

[0024] User preference data collection module: includes a user interface submodule and a data acquisition submodule. The user interface submodule provides a user preference setting interface; the data acquisition submodule is used to obtain historical data, geographic location information, and time information of users manually adjusting color temperature, and store the data in a local database;

[0025] Color temperature adjustment model training module: It has a data preprocessing unit, a model training unit, and a model evaluation unit. The data preprocessing unit cleans and normalizes the input data; the model training unit uses the back propagation algorithm and stochastic gradient descent method to train the deep learning model.

[0026] Real-time color temperature calculation module: This module inputs environmental data, image feature data, and user preference data into a trained model and calculates the appropriate LCD color temperature value for the current environment based on a preset formula. This module also includes a result verification unit to verify the calculation results.

[0027] Color temperature adjustment execution module: It consists of a driver chip and a control circuit. The driver chip adjusts the luminous intensity ratio of the three primary colors of red, green, and blue in the LCD backlight according to the received color temperature value; the control circuit is responsible for accurately transmitting the color temperature value to the driver chip;

[0028] Feedback optimization module: includes a feedback data collection unit and a model optimization unit. The feedback data collection unit collects user feedback data on the current color temperature; the model optimization unit uses the feedback data and the original input data as new training samples to retrain the color temperature adjustment model.

[0029] Furthermore, it also includes an abnormal situation processing module, which monitors the operating status of the environmental data acquisition module and the image feature extraction module in real time. When an abnormal situation is detected, it automatically switches to the default color temperature adjustment mode and displays a prompt message on the mobile phone interface.

[0030] Furthermore, it also includes an energy-saving mode adaptation module, which monitors the battery level of the mobile phone in real time. When the battery level of the mobile phone is lower than 20%, it automatically reduces the brightness of the LCD screen and adjusts the threshold range of the color temperature adjustment.

[0031] Compared with the existing technology, the beneficial effects of the present invention are:

[0032] In terms of environmental adaptability, the phone uses ambient light sensors and cameras to collect light intensity and image data, and combines this with AI algorithms to analyze factors such as the scene and object color, accurately matching the most suitable color temperature for different environments. Whether in bright sunlight outdoors, dimly lit indoors, or even in a colorful exhibition setting, the phone screen can quickly adjust the color temperature to present a clear, natural, and comfortable image, avoiding display distortion caused by environmental changes.

[0033] In terms of personalized service, the system collects historical data on users' manual color temperature adjustments, scene dwell time, ratings, and other preference information. Using AI algorithms to learn user habits, the system customizes a unique color temperature adjustment plan for each user. As usage increases, the system increasingly understands user preferences and automatically adjusts the screen color temperature to the user's satisfaction without manual intervention, greatly improving user convenience and comfort.

[0034] This application also features exception handling and energy-saving mode adaptation. When sensor data is abnormal, it automatically switches to the default mode to ensure usable operation. When the phone's battery is low, it rationally adjusts the color temperature adjustment strategy and screen brightness, ensuring a certain display quality while reducing energy consumption and extending the phone's battery life. In short, this technology comprehensively optimizes the display quality of mobile phone LCD screens through intelligent and personalized design, providing users with a higher-quality, smarter user experience and providing a new direction for the development of mobile phone display technology. BRIEF DESCRIPTION OF THE DRAWINGS

[0035] Figure 1 This is a schematic block diagram of the AI-based dynamic color temperature adjustment method for mobile phone LCD screens proposed in the present invention;

[0036] Figure 2 This is a schematic block diagram of the AI-based dynamic color temperature adjustment system for mobile phone LCD screens proposed by the present invention;

[0037] Figure 3 This is a bar chart comparing the color temperature adjustment accuracy of the traditional adjustment method and this adjustment method in different environments;

[0038] Figure 4 A line graph showing changes in user satisfaction with this adjustment method over time;

[0039] Figure 5 The bar chart shows the energy consumption ratio of different adjustment methods. DETAILED DESCRIPTION

[0040] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. 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 making creative efforts are within the scope of protection of the present invention.

[0041] In the description of the present invention, it should be understood that the terms "center", "longitudinal", "lateral", "length", "width", "thickness", "up", "down", "front", "back", "left", "right", "vertical", "horizontal", "top", "bottom", "inside", "outside", "clockwise", "counterclockwise" and the like to indicate orientations or positional relationships based on the orientations or positional relationships shown in the accompanying drawings, and are only 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 should not be understood as limiting the present invention.

[0042] In addition, the terms "first" and "second" are used for descriptive purposes only and cannot be understood as indicating or implying relative importance or implicitly indicating the number of technical features indicated. Therefore, the features defined as "first" and "second" may explicitly or implicitly include one or more of the said features. In the description of the present invention, the meaning of "multiple" is two or more, unless otherwise clearly and specifically defined. In addition, the terms "installed", "connected" and "connected" should be understood in a broad sense. For example, it can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection or an electrical connection; it can be a direct connection, or it can be indirectly connected through an intermediate medium, or it can be a connection between the two elements. For ordinary technicians in this field, the specific meanings of the above terms in the present invention can be understood according to specific circumstances. The present invention will be further described in detail below with reference to the accompanying drawings.

[0043] Reference Figures 1 to 5 :A method for dynamic color temperature adjustment of mobile phone LCD screen based on AI, comprising the following steps:

[0044] Environmental data collection steps: The phone is equipped with a BH1750 digital ambient light sensor, which offers excellent stability and high precision. Its measurement range is 1-100,000 lux, with an accuracy of ±2%, and it can continuously collect ambient light intensity data twice per second. It also features a rear-mounted 48-megapixel camera with autofocus and optical image stabilization, capturing ambient image data at 30 frames per second and a standard resolution of 4000×3000 pixels. The collected data is first stored in a data buffer area in the phone's memory. This buffer uses a circular queue data structure and can store light intensity data collected within the last 10 seconds and 100 frames of image data, preventing data loss during transmission and processing. When new data is written, the oldest data is automatically overwritten to ensure real-time and continuous data.

[0045] Image feature extraction step: The pre-processed image is input into a pre-trained convolutional neural network (CNN) model. The model structure consists of 15 convolution layers with kernel sizes of 3×3 and 5×5, which are used to extract local features of the image; 6 pooling layers, which use the maximum pooling method to reduce the data dimension and retain important features; and 4 fully connected layers, which integrate and classify the extracted features. (where x is the input image data, f j is the feature vector output by the model, C is the total number of categories, y i It can classify and predict object types (for the i-th category) and recognize more than 200 object types in images, such as natural landscapes, indoor furniture, electronic devices, etc. At the same time, it can extract feature information such as image color distribution and scene type, for example, to determine whether it is a bright outdoor scene, a dim bedroom scene, or a colorful exhibition scene.

[0046] User preference data collection steps: A dedicated user preference interface is designed within the display settings application of the mobile operating system. When the user manually adjusts the LCD screen's color temperature, the system records data such as the adjustment time, the color temperature value before and after adjustment, and the duration of the adjustment operation. Simultaneously, the phone's GPS positioning module is combined with network positioning technology to obtain geographic location information accurate to the street level; the system clock is used to obtain time information accurate to the second.

[0047] Color temperature adjustment model training steps: The ambient light intensity data, image feature data, and user preference data are integrated and cleaned as input data. The color temperature value that the user manually adjusts for more than 5 minutes and does not adjust again during this period is used as the output data of the user's satisfactory color temperature value. The deep learning model is trained using the backpropagation algorithm and the stochastic gradient descent method. The learning rate is initially set to 0.001 and is dynamically adjusted according to the training situation. The loss function uses the mean square error (MSE) formula (where n is the number of samples, y i is the true color temperature value, To predict the color temperature value), the loss function value is gradually reduced by continuously adjusting the weight and bias parameters in the model.

[0048] Real-time color temperature calculation steps: The mobile phone system inputs real-time ambient light intensity data, extracted image feature data, and user preference data into the trained color temperature adjustment model once per second. The model calculates the appropriate LCD screen color temperature value for the current environment according to the formula T = α × O + β × F + γ × P (where T is the predicted color temperature value, I is the ambient light intensity data, F is the image feature data vector, P is the user preference data vector, and α, β, and γ are weight coefficients obtained through training. For example, in one scenario, α = 0.3, β = 0.5, and γ = 0.2).

[0049] Color temperature adjustment steps: The calculated color temperature value is passed through the phone's I 2 C communication protocol, sent to the driver chip of the mobile phone's LCD screen at a frequency of 10 times per second. The driver chip integrates a high-precision PWM (pulse width modulation) controller, which can accurately adjust the luminous intensity ratio of the red, green, and blue primary color LEDs in the LCD screen backlight according to the received color temperature value. When the calculated color temperature value is high (such as exceeding 7500K), the driver chip will increase the PWM duty cycle of the blue light LED and reduce the PWM duty cycle of the red and green light LEDs accordingly. When the color temperature value is low (such as below 4500K), the PWM duty cycle of the red and green light LEDs is increased, and the PWM duty cycle of the blue light LED is reduced. During the adjustment process, the driver chip also adaptively adjusts the adjustment range of the three primary colors according to the current brightness level of the LCD screen to ensure that the screen brightness remains stable when the color temperature changes.

[0050] Feedback optimization step: After color temperature adjustment is performed, user feedback on the current color temperature is collected, including usage duration and whether manual adjustment was made. This module provides four rating options: "Very Satisfied," "Satisfied," "Average," and "Unsatisfied." After the user clicks "Rate," the system records the rating time, rating option, and current environmental and color temperature settings. This feedback data, along with the original input data, is used as new training samples to retrain the color temperature adjustment model, continuously optimizing the model's accuracy and user satisfaction.

[0051] The present invention also includes: abnormal situation processing steps: the system monitors the rate of change of the light intensity data collected by the ambient light sensor and the camera image parsing status in real time. When the light intensity data collected by the ambient light sensor changes by more than 10,000 lux within 0.2 seconds, or when 5 consecutive frames of image data collected by the camera cannot be normally parsed by the image decoding algorithm, the abnormal situation processing mechanism is triggered. The system automatically switches to the default color temperature adjustment mode, sets the color temperature of the LCD screen to 6700K, and displays a yellow exclamation mark warning icon in the status bar at the top of the mobile phone screen. Click the icon and a detailed prompt box will pop up, informing the user that "the current environmental data is abnormal and has been switched to the default color temperature mode." At the same time, the system records information such as the time when the abnormality occurred and environmental data in the log file to facilitate subsequent analysis and problem troubleshooting.

[0052] The present invention also includes: an energy-saving mode adaptation step: the mobile phone system monitors the battery power status in real time, and automatically starts the energy-saving mode when the power is lower than 15%. In energy-saving mode, the system first reduces the brightness of the LCD screen to 50% of the original, and adjusts the color temperature adjustment strategy at the same time. The lower limit of color temperature adjustment is increased from 4000K to 5000K, and the upper limit is increased from 8000K to 9000K. In this mode, the system reduces the frequency of environmental data collection, the ambient light sensor collection frequency is reduced to 0.5 times per second, and the camera only collects image data when the screen is unlocked or the application is called. At the same time, the calculation accuracy of the color temperature adjustment model is optimized to reduce the calculation complexity, further reduce system power consumption, and extend the battery life of the mobile phone.

[0053] In the present invention, in the image feature extraction step, the collected environmental image is preprocessed, including grayscale processing, noise reduction processing and normalization processing. Grayscale processing adopts the weighted average method, and the formula is Gray = 0.299×R+0.587×G+0.114×B (where R, G, and B are the red, green, and blue channel values ​​of the image, respectively), converting the color image into a grayscale image, reducing the amount of data while retaining the key information of the image. Noise reduction processing uses a median filtering algorithm, using a 3×3 filtering window to sort the pixel values ​​of each pixel point and its neighborhood, and taking the median as the new value of the pixel point, effectively removing interference such as salt and pepper noise in the image. Normalization processing maps the image pixel values ​​to the range of [0,1] to facilitate subsequent model processing. The formula is Where x is the original pixel value, x min with x max are the minimum and maximum pixel values ​​in the image, respectively.

[0054] In the present invention, in the step of collecting user preference data, the collaborative filtering algorithm is used to analyze the similarity between users. The cosine similarity formula between users is calculated (where I uv is the set of items that users u and v have jointly evaluated, r ui and r vi The cosine similarity between users is calculated based on the ratings given by users u and v for item i. In this scenario, the items are different color temperature settings, and the ratings represent the user's satisfaction with the different color temperatures. If a user's color temperature preferences differ from those of highly similar users in the same scenario, the system will push the preferences of similar users to that user for reference and selection. The system will also record user feedback to further improve user preference data.

[0055] In this paper, a transfer learning approach is employed during the color temperature adjustment model training step to improve training efficiency. A ResNet-50 model pre-trained on the large-scale public image dataset ImageNet is used as a foundation, retaining its convolutional layers and replacing and retraining the final fully connected layer. During training, the fully connected layer is first trained with a low learning rate. After initial model convergence, the entire model is fine-tuned, significantly reducing training time and data requirements. After 100 rounds of iterative training, the model's prediction accuracy on the test set is improved.

[0056] In this invention, a reinforcement learning algorithm is introduced in the feedback optimization step. The user's feedback is used as a reward signal, and the long-term cumulative reward is maximized by adjusting the model parameters. The Q-learning algorithm is used, and its core formula is Q(s, a)←Q(s, a)+α[r+γmax a′Q(s′, a′)-Q(s, a)] (where s is the current state, a is the current action, r is the immediate reward, γ is the discount factor, α is the learning rate, and s' is the next state). By continuously collecting user feedback, adjusting model parameters, and maximizing long-term cumulative rewards, the model can better adapt to the color temperature requirements of different users in various scenarios and optimize the color temperature adjustment strategy.

[0057] The present invention includes the following modules:

[0058] Environmental Data Acquisition Module: This module consists of an ambient light sensor and a camera. The ambient light sensor collects real-time ambient light intensity data, while the camera collects environmental image data. This module also includes a data cache unit that caches data collected within the last 10 seconds to prevent data loss.

[0059] Image Feature Extraction Module: This module includes an image preprocessing submodule and a feature extraction submodule. The image preprocessing submodule performs grayscale conversion, noise reduction, and normalization on the collected environmental images. The feature extraction submodule uses a convolutional neural network model to extract feature information such as object type, color distribution, and scene type from the image.

[0060] The user preference data collection module includes a user interface submodule and a data acquisition submodule. The user interface submodule provides a user preference setting interface; the data acquisition submodule is used to obtain historical data of users manually adjusting color temperature, including geographic location information, time information, etc., and store the data in a local database.

[0061] Color temperature adjustment model training module: This module includes a data preprocessing unit, a model training unit, and a model evaluation unit. The data preprocessing unit cleans and normalizes the input data; the model training unit uses the backpropagation algorithm and stochastic gradient descent to train the deep learning model; and the model evaluation unit evaluates model performance based on metrics such as mean square error.

[0062] Real-time color temperature calculation module: This module inputs environmental data, image feature data, and user preference data into a trained model and calculates the appropriate LCD color temperature for the current environment based on a preset formula. This module also includes a result verification unit to verify the rationality of the calculation results.

[0063] The color temperature adjustment execution module consists of a driver chip and a control circuit. The driver chip adjusts the luminous intensity ratio of the three primary colors of red, green, and blue in the LCD backlight according to the received color temperature value; the control circuit is responsible for accurately transmitting the color temperature value to the driver chip.

[0064] Feedback Optimization Module: This module includes a feedback data collection unit and a model optimization unit. The feedback data collection unit collects user feedback on the current color temperature. The model optimization unit uses the feedback data and the original input data as new training samples to retrain the color temperature adjustment model.

[0065] The present invention also includes an abnormality processing module that monitors the operating status of the environmental data acquisition module and the image feature extraction module in real time. When an abnormality is detected, it automatically switches to the default color temperature adjustment mode and displays a prompt message on the mobile phone interface.

[0066] The present invention also includes an energy-saving mode adaptation module that monitors the phone's battery level in real time. When the phone's battery level drops below 20%, it automatically reduces the brightness of the LCD screen and adjusts the color temperature threshold to reduce backlight energy consumption.

[0067] The above are only preferred specific embodiments of the present invention, but the scope of protection of the present invention is not limited thereto. Any technician familiar with this technical field, within the technical scope disclosed by the present invention, who makes equivalent replacements or changes based on the technical solutions and inventive concepts of the present invention, should be covered by the scope of protection of the present invention.

Claims

1. A method for dynamic color temperature adjustment of mobile phone LCD screen based on AI, characterized in that: The following steps are involved: Environmental data collection steps: The ambient light intensity data is collected through the mobile phone's ambient light sensor, and the camera collects environmental image data. The ambient light sensor uses a silicon photodiode sensor, and the camera supports autofocus and white balance functions; Image feature extraction step: Input the collected environmental image data into the trained convolutional neural network model, extract image feature information through the model, and use the Softmax function Classify and predict object type, x is the input image data, f j is the feature vector, C is the total number of categories, y i is the i-th category; User preference data collection steps: Set the user preference interface through the mobile phone application, collect user history data, user stay time data, and user color temperature rating data, and at the same time, obtain the geographic location and time information in the mobile phone system; Color temperature adjustment model training steps: ambient light intensity data, image feature data, and user preference data are used as input, and the color temperature value that the user is satisfied with is used as output. The back propagation algorithm and stochastic gradient descent method are used to train the deep learning model, and the loss function uses the mean square error formula. Adjust the model parameters, n is the number of samples, y i is the true color temperature value, To predict the color temperature value; Real-time color temperature calculation steps: Input the ambient light intensity data, extracted image feature data, and user preference data into the trained color temperature adjustment model. The LCD screen color temperature value is calculated using the formula T = α × I + β × F + γ × P, where T is the predicted color temperature value, I is the ambient light intensity data, F is the image feature data vector, P is the user preference data vector, and α, β, and γ are weight coefficients. Color temperature adjustment execution steps: send the color temperature value to the driver chip of the mobile phone LCD screen, and the driver chip adjusts the luminous intensity ratio of the three primary colors in the LCD screen backlight; Feedback optimization step: Collect user feedback data, use the feedback data and original input data as new training samples, and retrain the color temperature adjustment model.

2. The method for dynamic color temperature adjustment of mobile phone LCD screen based on AI according to claim 1, characterized in that: It also includes the steps for handling abnormal situations: when the light intensity data collected by the ambient light sensor fluctuates abnormally or the image data collected by the camera cannot be parsed normally, the system automatically switches to the default color temperature adjustment mode, the default color temperature value is 6500K, and a prompt message is displayed on the mobile phone interface.

3. The method for dynamic color temperature adjustment of mobile phone LCD screen based on AI according to claim 1, characterized in that: It also includes the energy-saving mode adaptation steps: when the mobile phone battery is less than 20%, the system automatically reduces the brightness of the LCD screen and increases the threshold range of color temperature adjustment while ensuring the user's visual comfort.

4. The method for dynamic color temperature adjustment of mobile phone LCD screen based on AI according to claim 1, characterized in that: In the image feature extraction step, the collected environmental image is preprocessed, including grayscale processing, noise reduction processing and normalization processing. The grayscale processing adopts the weighted average method with the formula Gray = 0.299×R+0.587×G+0.114×B, where R, G, and B are the red, green, and blue channel values ​​of the image, respectively. The noise reduction processing adopts the median filtering algorithm with a filter window size of 3×3; the normalization processing normalizes the image pixel values ​​to the range of [0,1].

5. The method for dynamic color temperature adjustment of mobile phone LCD screen based on AI according to claim 1, characterized in that: In the user preference data collection step, the collaborative filtering algorithm is used to analyze the similarity between users, and the cosine similarity formula between users is calculated. Recommend color temperature settings with similar user preferences to users, where I uv is the set of items that users u and v have jointly evaluated, r ui and r vi are the ratings of user u and user v on item i respectively.

6. The method for dynamic color temperature adjustment of mobile phone LCD screen based on AI according to claim 1, characterized in that: In the color temperature adjustment model training step, the transfer learning method is used. The CNN model pre-trained on a large-scale public image dataset is used as the basis. On this basis, it is fine-tuned for the mobile phone LCD screen color temperature adjustment task, reducing training time and data requirements.

7. The method for dynamic color temperature adjustment of mobile phone LCD screen based on AI according to claim 1, characterized in that: In the feedback optimization step, the reinforcement learning algorithm is introduced, and the user's feedback is used as the reward signal. The long-term cumulative reward is maximized by adjusting the model parameters. The Q-learning algorithm is used, and its core formula is Q(s,a)←Q(s,a)+α[r+γmax a′ Q(s′,a′)-Q(s,a)], where s is the current state, a is the current action, r is the immediate reward, γ is the discount factor, α is the learning rate, and s' is the next state, continuously optimizing the color temperature adjustment strategy.

8. A system using the AI-based dynamic color temperature adjustment method for mobile phone LCD screens according to any one of claims 1 to 5, characterized in that: Includes the following modules: Environmental data acquisition module: It consists of an ambient light sensor and a camera. The ambient light sensor is used to collect ambient light intensity data in real time; the camera is used to collect environmental image data. The module also includes a data cache unit to cache the data collected in the last 10 seconds to prevent data loss. Image feature extraction module: includes image preprocessing submodule and feature extraction submodule. The image preprocessing submodule is used to grayscale, reduce noise and normalize the collected environmental image. The feature extraction submodule uses a convolutional neural network model to extract feature information from images; User preference data collection module: includes a user interface submodule and a data acquisition submodule. The user interface submodule provides a user preference setting interface. The data acquisition submodule is used to obtain historical data, geographic location information, and time information of the user's manual adjustment of color temperature, and store the data in a local database; Color temperature adjustment model training module: It has a data preprocessing unit, a model training unit, and a model evaluation unit. The data preprocessing unit cleans and normalizes the input data. The model training unit uses the back propagation algorithm and stochastic gradient descent method to train the deep learning model; Real-time color temperature calculation module: This module inputs environmental data, image feature data, and user preference data into a trained model and calculates the appropriate LCD color temperature value for the current environment based on a preset formula. This module also includes a result verification unit to verify the calculation results. Color temperature adjustment execution module: It consists of a driver chip and a control circuit. The driver chip adjusts the luminous intensity ratio of the three primary colors of red, green, and blue in the LCD backlight according to the received color temperature value; the control circuit is responsible for accurately transmitting the color temperature value to the driver chip; Feedback optimization module: including a feedback data collection unit and a model optimization unit. The feedback data collection unit collects user feedback data on the current color temperature. The model optimization unit uses the feedback data together with the original input data as new training samples to retrain the color temperature adjustment model.

9. The system for the method for dynamic color temperature adjustment of mobile phone LCD screen based on AI according to claim 8, characterized in that: It also includes an abnormal situation processing module, which monitors the operating status of the environmental data acquisition module and the image feature extraction module in real time. When an abnormal situation is detected, it automatically switches to the default color temperature adjustment mode and displays a prompt message on the mobile phone interface.

10. The system for the method for dynamic color temperature adjustment of mobile phone LCD screen based on AI according to claim 8, characterized in that: It also includes an energy-saving mode adaptation module, which monitors the battery level of the mobile phone in real time. When the battery level of the mobile phone is less than 20%, it automatically reduces the brightness of the LCD screen and adjusts the threshold range of color temperature adjustment.

Images