Image display quality optimization method for display equipment

Abstract

The invention relates to the technical field of image display optimization, and particularly discloses an image display quality optimization method for display equipment, which is characterized in that multi-dimensional joint perceptual analysis is carried out on ambient light data and inherent characteristic parameters of a display to quantify the visual sensitivity of human eyes and equipment performance constraints; and generating a perceived quality reduction index representing the display quality attenuation degree. Then, further combining the perception quality reduction index and the key content characteristics of the original image frame data to be displayed to generate an image display optimization target parameter giving consideration to both visual comfort and content expressive force, and ensuring that the key points of the image are highlighted while the perception loss is compensated; and on this basis, a screen color temperature adjustment value, an optimal global brightness adjustment value and a local adjustment parameter are calculated so as to carry out pixel-level display optimization processing on original image frame data. According to the method, the image display quality under different ambient light conditions and different display conditions can be effectively improved, and better visual experience is provided for users.

Description

Technical Field

[0001] The present application relates to the technical field of image display optimization, and more specifically, to a method for optimizing image display quality of display equipment. Background Art

[0002] With the widespread use of various display devices in daily life, work, entertainment and other fields, users' requirements for image display quality are increasing. However, the ambient light conditions in which display devices are located are complex and changeable. From bright outdoors to dim indoors, factors such as the intensity and color temperature of the ambient light can have a significant impact on the image quality perceived by the human eye. For example, in a strong light environment, the screen content may appear dim, details may be lost, and colors may be distorted; while in a dark environment, the screen may be too bright and dazzling, which also affects viewing comfort and color accuracy. In addition, the inherent characteristic parameters of different displays, such as brightness range, contrast, color gamut, response time, etc., also directly restrict their optimal performance under specific ambient light conditions. Therefore, adaptive optimization of the display effect based on dynamically changing ambient light conditions and the characteristics of the display itself is a key technology to improve image display quality and enhance user experience.

[0003] Currently, existing methods for optimizing image quality on display devices rely on users manually adjusting parameters such as brightness, contrast, and color temperature. This approach is not only cumbersome but also difficult to achieve optimal results, especially for non-professional users. Some devices have introduced automatic brightness adjustment features based on ambient light sensors, which can mitigate the impact of ambient light variations to a certain extent. However, these methods are generally simple, focusing primarily on adjusting global brightness. They fail to fully consider the complex impact of ambient light on color perception and contrast details, and rarely integrate comprehensive optimization with the inherent characteristics of the display itself, resulting in significant room for improvement in display quality.

[0004] Therefore, a method for optimizing image display quality of display equipment is desired. Summary of the Invention

[0005] In order to solve the above technical problems, the present application is proposed. An embodiment of the present application provides an image display quality optimization method for display equipment, which performs a multi-dimensional joint perception analysis on ambient light data and inherent characteristic parameters of the display to quantify the visual sensitivity of the human eye and the performance constraints of the device, and generates a perception quality degradation index that characterizes the degree of display quality attenuation. Then, the perception quality degradation index and the key content features of the original image frame data to be displayed are further combined to generate image display optimization target parameters that take into account both visual comfort and content expressiveness, ensuring that the image focus is highlighted while compensating for the perception loss, and on this basis, the screen color temperature adjustment value, the optimal global brightness adjustment value and the local adjustment parameters are calculated to perform pixel-level display optimization processing on the original image frame data. This method can effectively improve the image display quality under different ambient light conditions and different display conditions, and provide users with a better visual experience.

[0006] According to one aspect of the present application, a method for optimizing image display quality of a display device is provided, comprising: Acquire ambient light data collected by the ambient light sensor; Obtaining original image frame data to be displayed and inherent characteristic parameters of the display; Extracting key content features from the original image frame data to be displayed; Inputting the ambient light data and the intrinsic characteristic parameters of the display into a perception state evaluation engine to obtain a perception quality degradation indicator; generating image display optimization target parameters based on the perceived quality degradation indicator and the key content features; Calculating a screen color temperature adjustment value, an optimal global brightness adjustment value, and a local adjustment parameter as display quality adjustment parameters based on the image display optimization target parameter and the key content feature; The original image frame data to be displayed is subjected to pixel-level processing based on the display quality adjustment parameter to obtain ambient light adaptive optimized image frame data.

[0007] Compared with the prior art, the image display quality optimization method for display equipment provided by this application performs a multi-dimensional joint perception analysis of ambient light data and inherent characteristic parameters of the display to quantify the visual sensitivity of the human eye and the performance constraints of the device, and generates a perceptual quality degradation index that characterizes the degree of display quality attenuation. Then, the perceptual quality degradation index and the key content features of the original image frame data to be displayed are further combined to generate image display optimization target parameters that take into account both visual comfort and content expressiveness, ensuring that the image focus is highlighted while compensating for perceptual losses. On this basis, the screen color temperature adjustment value, the optimal global brightness adjustment value and the local adjustment parameters are calculated to perform pixel-level display optimization processing on the original image frame data. This method can effectively improve the image display quality under different ambient light conditions and different display conditions, providing users with a better visual experience. BRIEF DESCRIPTION OF THE DRAWINGS

[0008] The above and other purposes, features, and advantages of the present application will become more apparent through a more detailed description of the embodiments of the present application in conjunction with the accompanying drawings. The accompanying drawings are intended to provide a further understanding of the embodiments of the present application and constitute a part of the specification. Together with the embodiments of the present application, they are used to explain the present application and do not constitute a limitation of the present application. In the drawings, the same reference numerals generally represent the same components or steps.

[0009] Figure 1 Flowchart of a method for optimizing image display quality of a display device according to an embodiment of the present application.

[0010] Figure 2 Schematic diagram of data flow of a method for optimizing image display quality of a display device according to an embodiment of the present application.

[0011] Figure 3 4 is a flowchart of sub-step S3 of the method for optimizing image display quality of display equipment according to an embodiment of the present application.

[0012] Figure 4 4 is a flowchart of sub-step S4 of the method for optimizing image display quality of display equipment according to an embodiment of the present application.

[0013] Figure 5 4 is a flowchart of sub-step S42 of the method for optimizing image display quality of display equipment according to an embodiment of the present application.

[0014] Figure 6 4 is a flowchart of sub-step S423 of the method for optimizing image display quality of display equipment according to an embodiment of the present application. DETAILED DESCRIPTION

[0015] As used in this application and the claims, unless the context clearly indicates otherwise, the words "a," "an," "an," and / or "the" are not intended to refer to the singular but may include the plural. Generally speaking, the terms "comprises" and "include" only indicate the inclusion of the steps and elements specifically identified, and these steps and elements do not constitute an exclusive list. A method or apparatus may also include other steps or elements.

[0016] Although the present application makes various references to certain modules in the system according to embodiments of the present application, any number of different modules can be used and run on the user terminal and / or server. The modules are illustrative only, and different aspects of the system and method can use different modules.

[0017] Flowcharts are used in this application to illustrate the operations performed by the systems according to the embodiments of the present application. It should be understood that the preceding or following operations are not necessarily performed in exact order. Instead, the various steps may be processed in reverse order or simultaneously, as needed. Furthermore, other operations may be added to these processes, or one or more operations may be removed from these processes.

[0018] Below, the exemplary embodiments according to the present application will be described in detail with reference to the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all the embodiments of the present application, and it should be understood that the present application is not limited to the exemplary embodiments described herein.

[0019] It is worth noting that in this application, all actions to obtain data are carried out in compliance with the relevant data protection laws and policies of the country where they are located and with the authorization given by the owner of the corresponding device.

[0020] In response to the technical problems described in the above background technology, this application proposes a method for optimizing image display quality for display equipment, which performs a multi-dimensional joint perception analysis of ambient light data and inherent characteristic parameters of the display to quantify the visual sensitivity of the human eye and the performance constraints of the device, and generates a perceptual quality degradation index that characterizes the degree of display quality attenuation. Then, the perceptual quality degradation index and the key content features of the original image frame data to be displayed are further combined to generate image display optimization target parameters that take into account both visual comfort and content expressiveness, ensuring that the image focus is highlighted while compensating for perceptual losses, and on this basis, the screen color temperature adjustment value, the optimal global brightness adjustment value and the local adjustment parameters are calculated to perform pixel-level display optimization processing on the original image frame data. This method can effectively improve the image display quality under different ambient light conditions and different display conditions, providing users with a better visual experience.

[0021] Figure 1Flowchart of a method for optimizing image display quality of a display device according to an embodiment of the present application. Figure 2 Schematic diagram of data flow for the method for optimizing image display quality of display equipment according to an embodiment of the present application. Figure 1 and Figure 2 As shown, the image display quality optimization method for display equipment includes the following steps: S1, obtaining ambient light data collected by an ambient light sensor; S2, obtaining original image frame data to be displayed and inherent characteristic parameters of the display; S3, extracting key content features from the original image frame data to be displayed; S4, inputting the ambient light data and the inherent characteristic parameters of the display into a perception state evaluation engine to obtain a perception quality degradation index; S5, generating image display optimization target parameters based on the perception quality degradation index and the key content features; S6, calculating a screen color temperature adjustment value, an optimal global brightness adjustment value and a local adjustment parameter as display quality adjustment parameters based on the image display optimization target parameters and the key content features; S7, performing pixel-level processing on the original image frame data to be displayed based on the display quality adjustment parameters to obtain ambient light adaptive optimized image frame data.

[0022] In the above-mentioned method for optimizing the image display quality of a display device, step S1 obtains ambient light data collected by an ambient light sensor. It should be understood that the ambient light conditions (e.g., brightness and color temperature) of the display device are the most direct and dynamically changing external factors that affect the human eye's perception of image quality. For example, screen content may be difficult to see under strong sunlight, while an overly bright screen in a dim room may be glaring. Therefore, in order to enable subsequent image display quality optimization to accurately adapt to the current actual viewing environment, this application, based on the principle of photoelectric sensing, obtains ambient light data by real-time monitoring and quantifying the surrounding lighting conditions through an ambient light sensor deployed on the display device. Specifically, the ambient light sensor (ALS) integrated into the display device (such as a smartphone, tablet, monitor, or car screen) measures the illuminance (e.g., in lux) and color temperature (e.g., in Kelvin) of the current environment periodically or when triggered by a specific event (e.g., screen lighting, application switching). The analog light signal collected by the ambient light sensor is converted into digital data by the ADC inside the sensor, and after preliminary filtering and calibration, it is provided to the subsequent processing module as structured data (e.g., a vector containing illuminance and color temperature values), laying an objective environmental input foundation for subsequent visual perception status evaluation and targeted image optimization.

[0023] In the actual implementation process, ambient light sensors are usually integrated into display devices, such as smartphones, tablets, monitors or car screens, to effectively monitor the surrounding ambient light conditions. These sensors are based on the principle of photoelectric sensing and convert the collected analog light signals into digital data, thereby providing structured input information for subsequent processing modules. Specifically, during the design phase of display equipment, engineers will carefully select suitable ambient light sensors based on the device's dimensions, usage scenarios, and expected functional requirements, and place them in locations on the device surface that are not easily obstructed, such as the frame or above the screen, to ensure that the sensor can accurately capture changes in external light sources.

[0024] When a user turns on a display device, the ambient light sensor begins measuring the illuminance and color temperature of the current environment periodically or when triggered by specific events (such as turning on the screen or switching applications). The illuminance value mentioned here is usually measured in lux, which describes the amount of luminous flux received per unit area; while the color temperature is measured in Kelvin (K), which reflects the color characteristics of the light source, that is, the trend from warm to cool tones. In actual operation, the analog-to-digital converter (ADC) inside the ambient light sensor converts the captured analog light signal into a digital data stream. After a series of preprocessing steps, including filtering and calibration, to eliminate noise interference and correct potential deviations, accurate and reliable ambient light data is finally generated.

[0025] It's worth mentioning that to further improve the accuracy and stability of ambient light data, modern display devices are often equipped with adaptive adjustment mechanisms supported by advanced algorithms. This means that in addition to basic light intensity and color temperature measurements, the system can also dynamically adjust the sensor's operating parameters, such as sampling frequency and sensitivity threshold, based on historical data and current environmental characteristics to better adapt to complex and changing application scenarios. In addition, considering that different types of display equipment may face very different lighting conditions, manufacturers will also develop customized correction models for specific product lines to ensure optimal perception, whether in bright sunlight or dim indoor environments.

[0026] Throughout the data acquisition process, ambient light sensors must not only respond to external light intensity but also consider the impact of various factors on the human visual experience. For example, in high-brightness environments, strong direct sunlight can cause screen content to appear dim and unclear, with significant loss of detail. In low-brightness environments, excessive screen brightness can cause visual fatigue and even damage eye health. Therefore, fine-tuning the operating mode of the ambient light sensor can effectively alleviate these problems.

[0027] In the above-mentioned method for optimizing image display quality for display equipment, step S2 involves obtaining the raw image frame data to be displayed and the inherent characteristic parameters of the display. It should be understood that this application considers that the final image display effect depends not only on the external ambient light but also on the characteristics of the content to be displayed and the physical performance limits of the display device itself (such as maximum brightness, contrast, and color gamut). Therefore, to comprehensively assess the potential and bottlenecks of the display system under current conditions and develop an optimization strategy that is both consistent with content expression and achievable within the device's capabilities, this application constructs a complete visual perception status assessment system by obtaining raw pixel data from the image source and retrieving key performance indicators of the display from a device information library. In the specific implementation process, the raw image frame data to be displayed is first obtained from the video decoder, graphics rendering pipeline, or static image library. Simultaneously, the inherent characteristic parameters of the display, such as maximum / minimum brightness (nits), native contrast, color gamut, and refresh rate, are read, either pre-stored or obtained through EDID (Extended Display Identification Data).

[0028] In the above-mentioned method for optimizing image display quality for display equipment, the step S3 extracts key content features from the original image frame data to be displayed. Specifically, due to the significant differences in the importance of different areas in the image (such as text that needs to have enhanced edges and portraits that need to preserve skin color), key information will be lost if a global unified optimization strategy is adopted. Therefore, in order to identify the content semantic priority of the original image frame data to be displayed and guide the local optimization strategy, this application further extracts key content features from the original image frame data to be displayed based on computer vision feature extraction technology to understand the visual content and structure of the original image frame data to be displayed, thereby providing more accurate guidance for subsequent display quality optimization. Among them, Figure 3 FIG. 1 is a flow chart of sub-step S3 of the method for optimizing image display quality of display equipment according to an embodiment of the present application. Figure 3 As shown, the step S3 includes the steps of: S31, performing semantic segmentation on the original image frame data to be displayed to obtain a semantic mask map of the original image to be displayed; S32, performing CIELAB color space conversion and local neighborhood color difference calculation on the original image frame data to be displayed to generate a color attention map of the original image to be displayed; S33, performing Luma channel conversion and local neighborhood contrast calculation on the original image frame data to be displayed to generate a contrast attention map of the original image to be displayed; S34, performing cascade processing along the channel dimension on the semantic mask map of the original image to be displayed, the color attention map of the original image to be displayed and the contrast attention map of the original image to be displayed to obtain the key content features.

[0029] Specifically, step S31 performs semantic segmentation on the original image frame data to be displayed to obtain a semantic mask of the original image to be displayed. Specifically, the original image frame data to be displayed is semantically annotated using an FCN (fully convolutional network) model to identify different objects (such as people, scenery, text, etc.) in the image and their locations, generating a semantic mask of the same size as the original image. The element value of each pixel represents the object category to which the pixel at that location belongs.

[0030] Specifically, step S32 performs CIELAB color space conversion and local neighborhood color difference calculation on the original image frame data to be displayed to generate a color attention map for the original image to be displayed. Specifically, the original image frame data to be displayed is converted from RGB space to CIELAB color space, and the color difference ΔE value between each pixel in the CIELAB image and its neighboring pixels (e.g., a 3×3 window) is calculated. The color attention map is then normalized to generate a color attention map. This color attention map reflects the degree of color change in a local area of the image and can be used to identify detailed information such as edges and textures in the image, facilitating the precise positioning of color-sensitive areas and guiding visual attention.

[0031] Specifically, step S33 performs Luma channel conversion and local neighborhood contrast calculation on the original image frame data to be displayed to generate a contrast attention map of the original image to be displayed. That is, the original image frame data to be displayed is converted to the Luma channel, and the brightness fluctuation intensity (variance) and edge sharpness (gradient) of each local area of the image are calculated through a sliding window to generate a local brightness variance map and a gradient amplitude map. The two are normalized and then weightedly fused to construct a contrast attention map to highlight areas in the image with high contrast that are easy to attract visual attention, and retain the contrast and details of important areas during the optimization process.

[0032] Specifically, in step S34, the semantic mask map of the original image to be displayed, the color attention map of the original image to be displayed, and the contrast attention map of the original image to be displayed are cascaded along the channel dimension to obtain the key content features. That is, the semantic mask map, the color attention map, and the contrast attention map are cascaded and fused to form multi-channel key content features. Based on this, the key content features not only contain the semantic information of different objects in the image, but also integrate the color distribution changes and contrast characteristics of the image, thereby providing comprehensive content guidance for subsequent image display optimization, ensuring that optimized resources give priority to the presentation of key content, and realizing content-sensitive intelligent display.

[0033] In the above-mentioned method for optimizing image display quality of display equipment, in step S4, the ambient light data and the inherent characteristic parameters of the display are input into a perception state evaluation engine to obtain a perception quality degradation index. It should be understood that the present application takes into account that the perception of image quality by the human eye is not simply determined independently by the ambient light or the characteristics of the display, but is a comprehensive result of the interaction between the two. Therefore, in order to quantitatively evaluate the degree of attenuation or distortion of the image quality actually perceived by the human eye compared to the ideal state under the current specific ambient light conditions and display performance constraints, the present application constructs a perception state evaluation engine by utilizing a deep learning algorithm, and performs a deep joint analysis of the ambient light data and the inherent characteristic parameters of the display to simulate and quantify the response characteristics of the human visual system to the "ambient light-display" combination state, thereby generating a perception quality degradation index that characterizes the degree of attenuation of image display quality. Among them, Figure 4 FIG. 4 is a flow chart of sub-step S4 of the method for optimizing image display quality of display equipment according to an embodiment of the present application. Figure 4 As shown, the step S4 includes the steps of: S41, respectively performing structured embedded coding on the ambient light data and the display inherent characteristic parameters to obtain an ambient light state structured embedded coding vector and a display inherent characteristic parameter structured embedded coding vector; S42, performing multi-scale progressive interactive perception on the ambient light state structured embedded coding vector and the display inherent characteristic parameter structured embedded coding vector to obtain an ambient light-display parameter multi-scale progressive interactive perception coding vector; S43, performing feature decoding on the ambient light-display parameter multi-scale progressive interactive perception coding vector to obtain the perception quality degradation indicator.

[0034] Specifically, in one specific example of the present application, step S41 includes normalizing the ambient light data and then performing embedding encoding based on a fully connected layer to obtain a structured embedding encoding vector for the ambient light state. It should be understood that since the original ambient light data (e.g., illuminance values such as 10,000 Lux and color temperature values such as 6,500K) and the display's intrinsic characteristic parameters (e.g., maximum brightness of 500 nits, contrast ratio of 1,000:1, and color gamut coverage of 90% sRGB) are typically heterogeneous, with discrete numerical values or categorical information of varying scales and physical meanings, directly using these raw, sparse, or multi-dimensional data for joint analysis results in low computational efficiency and difficulty capturing the interrelationships between parameters, making it impossible to fully understand the true impact of each parameter. Therefore, in order to convert the multi-source, heterogeneous raw input data into a unified mathematical representation that is easier for computers to understand and process, the present application, based on the principle of representation learning, further constructs a data encoding network to perform independent structured embedding encoding on the ambient light data and the display's intrinsic characteristic parameters. Specifically, the numerical data (illuminance and color temperature) of the ambient light data is first normalized and then input into a fully connected layer for preliminary feature extraction. This layer learns the underlying relationship between illuminance and color temperature. A ReLU activation function is then used to introduce nonlinear characteristics to enhance feature expression, resulting in a structured embedded coding vector for the ambient light state. Similarly, for the display's intrinsic characteristic parameters, normalization and fully connected layer feature extraction are used to learn the display's performance under the combined influence of multiple source parameters. A ReLU activation function is then used to enhance the nonlinear expression of the features, resulting in a structured embedded coding vector for the display's intrinsic characteristic parameters.

[0035] Specifically, step S42 performs multi-scale progressive interactive perception on the ambient light state structured embedded coding vector and the display intrinsic characteristic parameter structured embedded coding vector to obtain an ambient light-display parameter multi-scale progressive interactive perception coding vector. Specifically, since the impact of ambient light conditions and display characteristics on the final perceived quality of the human eye is not a simple linear superposition, but rather a nonlinear interaction at different levels of abstraction—for example, the ambient light intensity may have a strong impact on the effective contrast of the display, and the extent of this impact may vary depending on the color temperature of the ambient light and the color gamut performance of the display—simply splicing and fusing the ambient light state structured embedded coding vector and the display intrinsic characteristic parameter structured embedded coding vector is difficult to fully capture this multifaceted and deep coupling relationship. Therefore, in order to deeply model and characterize the interdependence and joint effects between the ambient light state and the display characteristics, and simulate the comprehensive response of the human visual system to visual stimuli under a specific "environment-device" combination, this application further constructs a multi-scale progressive interactive perception network to perform multi-level interactive processing on the structured embedded coding vector of the ambient light state and the structured embedded coding vector of the display inherent characteristic parameters, so as to gradually reveal and integrate the joint action mechanism of the two from microscopic details to overall macroscopic characteristics, and finally generate a multi-scale progressive interactive perception coding vector of ambient light-display parameters. Among them, Figure 5 FIG. 4 is a flow chart of sub-step S42 of the method for optimizing image display quality of display equipment according to an embodiment of the present application. Figure 5 As shown, the step S42 includes the steps of: S421, performing low-level feature interaction encoding based on a multi-layer perceptron model on the structured embedded coding vector of the ambient light state and the structured embedded coding vector of the intrinsic characteristic parameters of the display to obtain a low-level joint perception coding vector of the ambient light-display parameters; S422, performing multi-level high-order feature interaction on the structured embedded coding vector of the ambient light state and the structured embedded coding vector of the intrinsic characteristic parameters of the display to obtain a middle-level joint perception coding vector of the ambient light-display parameters and a deep-level joint perception coding vector of the ambient light-display parameters; S423, performing multi-scale complementary progressive joint perception on the low-level joint perception coding vector of the ambient light-display parameters, the middle-level joint perception coding vector of the ambient light-display parameters and the deep-level joint perception coding vector of the ambient light-display parameters to obtain a multi-scale progressive interaction perception coding vector of the ambient light-display parameters.

[0036] More specifically, step S421 is expressed as follows: in, Indicates adding by position point, Represents the structured embedding encoding vector of the ambient light state, Represents the structured embedded coding vector of the display’s intrinsic characteristic parameters, represents the multi-layer perceptron model, Represents the low-level joint perceptual encoding vector of ambient light and display parameters.

[0037] That is, low-level feature interaction encoding is performed through a multi-layer perceptron model, the basic features of ambient light and display inherent characteristic parameters are preliminarily integrated, the input ambient light state structured embedding coding vector and the display inherent characteristic parameter structured embedding coding vector are processed layer by layer, and the low-level basic features of the two parameters are mined and extracted through the connection weights and activation functions between neurons, and these features are interactively fused to finally output the ambient light-display parameter low-level joint perception coding vector, so as to effectively reduce the complexity of the original parameter information, establish a preliminary connection between the ambient light and display inherent characteristic parameters in the low-level dimension, and improve the efficiency and accuracy of subsequent feature interaction and comprehensive analysis.

[0038] More specifically, step S422 includes: first, performing multi-level implicit feature extraction on the ambient light state structured embedded coding vector and the display intrinsic characteristic parameter structured embedded coding vector to obtain a middle-level implicit coding vector of the ambient light state feature, a middle-level implicit coding vector of the display intrinsic characteristic feature, a deep-level implicit coding vector of the ambient light state feature, and a deep-level implicit coding vector of the display intrinsic characteristic feature, which can be expressed as follows: in, and Represent the weight matrix and bias term of the middle-level implicit feature extraction network, represents the ReLU activation function, Represents the mid-level implicit coding vector of the ambient light state feature, represents the implicit coding vector of the inherent characteristics of the display, and Represent the weight matrix and bias term of the deep implicit feature extraction network, represents the Sigmoid activation function, and They represent the deep implicit coding vector of the ambient light state feature and the deep implicit coding vector of the display inherent characteristic feature respectively.

[0039] That is, by performing multi-level implicit feature extraction on the structured embedded coding vector of the ambient light state and the structured embedded coding vector of the display inherent characteristic parameters, the key information of the two types of data is captured from different abstraction levels, so that the obtained middle-level implicit coding vector of the ambient light state feature and the middle-level implicit coding vector of the display inherent characteristic feature can extract structured information with certain invariance in the ambient light state, such as the regular change pattern of factors such as ambient light intensity and color temperature, and the structured features of parameters such as brightness range and color gamut in the inherent characteristics of the display. The deep-level implicit coding vector of the ambient light state feature and the deep-level implicit coding vector of the display inherent characteristic feature can capture the global impact trend of the ambient light state and the core constraint information of the inherent characteristics of the display. After high abstraction and information compression, a global semantic representation closely related to the display quality assessment task is formed, providing high-level information focusing on core concepts for deep-level feature interaction.

[0040] Then, a mid-level feature interaction based on the cross-attention mechanism is performed on the mid-level implicit coding vector of the ambient light state feature and the mid-level implicit coding vector of the display intrinsic characteristic feature to obtain the mid-level joint perceptual coding vector of the ambient light-display parameter, which is expressed as: in, For the attention fusion network, Indicates point multiplication by position, represents the mid-level joint perceptual encoding vector of ambient light-display parameters, for The characteristic scale value of represents the transpose of a vector, is the normalized exponential function.

[0041] Specifically, a cross-attention mechanism is used to interact the mid-level implicit encoding vectors of the ambient light state features with the mid-level implicit encoding vectors of the display's inherent characteristics, thereby establishing a precise correlation mapping between the two at the structured information level. This interaction not only avoids data noise interference, but also, while retaining the core structured information of ambient light and device characteristics, explores the fine-grained synergy between the two in the mid-level semantic space. The resulting mid-level joint perception encoding vector of ambient light and display parameters combines the scene-specificity of ambient light influences with the constraints of display characteristics, providing a connecting mid-level interactive feature for subsequent multi-scale feature fusion.

[0042] Finally, a deep-level feature interaction based on linear projection gated interaction is performed on the deep implicit coding vector of the ambient light state feature and the deep implicit coding vector of the display intrinsic characteristic feature to obtain the deep-level joint perception coding vector of the ambient light and display parameters, which is expressed as follows: in, and are different low-rank projection matrices, represents the GELU activation function, Represents the LayerNorm normalization function, is the weight matrix of the deep level feature fusion network, represents the ambient light-display parameter feature projection gated interaction encoding vector, Represents the deep-level joint perceptual encoding vector of ambient light and display parameters.

[0043] Specifically, the deep implicit coding vectors of ambient light state features and the deep implicit coding vectors of the display's inherent characteristics are processed through linear projection gated interaction, capturing the core relationship between ambient light and display characteristics at a highly abstract semantic level. This interactive approach preserves the core semantic representations of ambient light and device characteristics while constructing a deep-level joint perceptual coding vector of ambient light and display parameters that focuses on the key factors of display quality degradation. This provides deeply fused features with global decision-making value for the precise quantification of perceptual quality degradation indicators, avoiding interference from low-level details.

[0044] Figure 6 FIG. 4 is a flowchart of sub-step S423 of the method for optimizing image display quality of display equipment according to an embodiment of the present application. Figure 6 As shown, the step S423 includes the steps of: S4231, performing complementary interactive fusion based on a gating mechanism on the low-level joint perception coding vector of the ambient light-display parameters and the medium-level joint perception coding vector of the ambient light-display parameters to obtain the low-level joint perception coding vector of the ambient light-display parameters; S4232, performing cross-level interaction based on a cross-attention mechanism on the low-level joint perception coding vector of the ambient light-display parameters and the deep-level joint perception coding vector of the ambient light-display parameters to obtain the multi-scale progressive interactive perception coding vector of the ambient light-display parameters.

[0045] In a specific example of the present application, step S4231 is expressed as follows: in, represents feature cascade, represents the weight matrix of the progressive complementary perception network, express and The gated interaction coefficient between Represents the low-level joint perceptual encoding vector of ambient light and display parameters.

[0046] It should be understood that this application uses a gating mechanism to perform complementary interactive fusion of the low-level joint perception coding vector of ambient light and display parameters and the mid-level joint perception coding vector of ambient light and display parameters. This can preserve the original detailed information of ambient light and display characteristics while incorporating the structured correlation features of the mid-level abstraction. In this way, it avoids noise interference from low-level data and compensates for key details that may be lost in the mid-level abstraction. The generated low-level and mid-level joint perception coding vector of ambient light and display parameters has both parameter accuracy and pattern relevance, providing a hierarchical and complementary feature representation for subsequent cross-level deep interaction.

[0047] In particular, considering that when hierarchical dynamic feature integration is performed on the low-level joint perception coding vector of the ambient light-display parameters, the middle-level joint perception coding vector of the ambient light-display parameters and the deep-level joint perception coding vector of the ambient light-display parameters, the low-level joint perception coding vector of the ambient light-display parameters and the middle-level joint perception coding vector of the ambient light-display parameters actually generate gated interaction coefficients, which are then fused based on the gated interaction coefficients. At the same time, the fused low-level joint perception coding vector of the ambient light-display parameters and the deep-level joint perception coding vector of the ambient light-display parameters are further subjected to cross-level interaction based on the converter architecture, that is, via the weight matrix 、 and The covariate tensor field is constrained to achieve feature interaction. Here, it is considered that the detuning of the constraint configuration of the covariate tensor field will lead to an interaction inhibition effect, thereby affecting the expression effect of the multi-scale progressive interaction perception coding vector of the ambient light-display parameter. Therefore, in a preferred example of the present application, the step S4232 includes: obtaining a query embedding matrix, a key embedding matrix and a value embedding matrix, and dynamically optimizing the weight distribution of the query embedding matrix, the key embedding matrix and the value embedding matrix based on multi-objective collaborative constraints to obtain an optimized query embedding matrix, an optimized key embedding matrix and an optimized value embedding matrix.

[0048] Specifically, first, for the pre-trained matrix 、 and , taking the combination of the three as the covariate tensor field, calculate the path integral representation of each decomposition field of the covariate tensor field: in, 、 and denote the query embedding matrix, key embedding matrix and value embedding matrix respectively, 、 and They represent the optimized query embedding matrix, optimized key embedding matrix, and optimized value embedding matrix respectively.

[0049] Then, construct the matrix balanced distribution alignment loss function: in, The spectral norm, which is the sum of the eigenvalues of the matrix, represents the scaling adjustment coefficient, Represents the matrix balanced distribution alignment loss function value.

[0050] In this way, the accumulation of incompleteness of the constraint connection of the covariate tensor field is suppressed by the path integral of each decomposition field of the covariate tensor field to avoid the formation of geometric phase superposition that hinders the interaction of local fields. Then, the stable state of the constraint configuration can be maintained by the tensor equilibrium equivalence represented by the spectral norm, thereby realizing the weight matrix 、 and The feature interaction is maintained, and the expression effect of the multi-scale progressive interaction perception coding vector of the ambient light-display parameter is improved.

[0051] Then, the optimized query embedding matrix is used to embed the low-level joint perceptual coding vector of the ambient light-display parameter to obtain a query vector, and the optimized key embedding matrix and the optimized value embedding matrix are used to embed the deep-level joint perceptual coding vector of the ambient light-display parameter to obtain a key vector and a value vector, respectively, which can be expressed as follows: in, 、 and Represent the query vector, key vector and value vector respectively.

[0052] Specifically, by optimizing the query embedding matrix, the key embedding matrix, and the value embedding matrix, the low- and medium-level joint perception encoding vectors of the ambient light and display parameters are embedded in the deep-level joint perception encoding vectors of the ambient light and display parameters, thereby constructing a semantic association space suitable for the display quality optimization task for cross-level feature interaction. This encoding method enables the query vector to specifically retrieve core semantic information strongly related to low- and medium-level details from the deep-level features. The key vector and value vector provide task-oriented optimized global constraints for this retrieval process, thereby achieving a precise match between fine-grained scene parameters and high-level semantic constraints in subsequent cross-level interactions.

[0053] Finally, a cross-level interaction based on a converter architecture is performed on the query vector, the key vector, and the value vector to obtain the multi-scale progressive interaction perception encoding vector of the ambient light-display parameter, which is expressed as: in, A multi-scale progressive interaction perceptual coding vector representing the ambient light-display parameter is provided.

[0054] Specifically, the query vector, key vector, and value vector interact across multiple levels through a transformer architecture, establishing a global association and dynamic weighting of ambient light and display characteristics in a multi-scale feature space. This interaction allows precise low-level parameter details to be reorganized and weighted within a high-level semantic framework, while high-level global constraints pinpoint key dimensions that require optimization at lower levels. The resulting multi-scale progressive interaction perception encoding vector for ambient light and display parameters combines a fine-grained description of the dynamic changes in ambient light with high-level semantic guidance for display characteristic constraints, providing a richly layered and semantically aligned feature representation for the subsequent quantitative analysis of perceptual quality degradation indicators.

[0055] Specifically, in a specific example of the present application, step S43 includes: inputting the multi-scale progressive interaction perceptual encoding vector of ambient light and display parameters into a feature decoder based on a multi-layer perceptron to obtain the perceptual quality degradation indicator. Specifically, in order to convert the multi-scale progressive interaction perceptual encoding vector of ambient light and display parameters from an abstract, high-dimensional internal feature representation into a specific and quantifiable perceptual quality degradation indicator, the present application designs a feature decoding network. The network, through a multi-layer perceptron (MLP) structure, performs multi-level feature learning and progressive decoding on the multi-scale progressive interaction perceptual encoding vector to restore the degree of attenuation of image quality perception by the human eye under different "ambient light-display" combinations. The network maps the multi-scale progressive interaction perceptual encoding vector of ambient light and display parameters to a predefined perceptual quality degradation indicator space, and outputs quantitative evaluation values of multiple dimensions such as contrast loss, brightness loss, and color saturation loss as perceptual quality degradation indicators, thereby accurately quantifying the degree of image display quality degradation and providing data support and reference basis for subsequent image display optimization.

[0056] In the above-mentioned method for optimizing image display quality for display equipment, step S5 generates image display optimization target parameters based on the perceived quality degradation indicator and the key content features. It should be understood that simply compensating for visual perception loss may result in a decrease in the content fidelity of the original image (e.g., global brightening at the expense of highlight detail). Therefore, to balance the visual comfort and content expressiveness of the original image frame data to be displayed, this application further solves for optimization target parameters that balance environmental adaptability and content fidelity based on the principles of multi-objective optimization game and constraint solving. Specifically, an objective function is first constructed to minimize the perceived quality degradation indicator while maximizing the weight for content feature restoration (e.g., assigning a higher edge sharpness restoration weight to text regions and a skin tone fidelity constraint to facial regions). A dynamic weight allocation algorithm is used to adjust optimization priorities based on semantic segmentation results, color attention distribution, and contrast attention distribution (e.g., prioritizing resources to reduce distortion in high-contrast or colorful image regions and giving higher contrast restoration weights to text regions than to image background regions). Subsequently, the Lagrange multiplier method is used to solve for the optimal parameter combination within the device performance boundaries (such as the maximum brightness limit and the color gamut boundaries), generating an optimized target parameter set (including the target global contrast gain coefficient, local sharpening intensity, color gamut expansion ratio, and brightness layering threshold). In this way, global brightness can be reduced in low-light scenes to alleviate visual fatigue, while enhancing text edge sharpness to maintain readability, or the color gamut range can be dynamically expanded to match the ambient light conditions when playing HDR content.

[0057] In the above-mentioned method for optimizing the image display quality of display equipment, the step S6 calculates the screen color temperature adjustment value, the optimal global brightness adjustment value and the local adjustment parameters as display quality adjustment parameters based on the image display optimization target parameters and the key content features. It should be understood that the image display optimization target parameters define the quantitative requirements of global and local optimization (such as contrast gain coefficient, color gamut expansion ratio), and the key content features clarify the priorities of different areas (such as text areas need to enhance sharpness, and portrait areas need to preserve skin color fidelity), and image display optimization needs to be achieved through the collaboration of hardware control and image processing. Therefore, in order to further map the two collaboratively into executable hardware control and image processing instructions. This application is based on the principles of multimodal parameter fusion and hierarchical control to calculate the screen color temperature adjustment value, the optimal global brightness adjustment value and the local adjustment parameters as display quality adjustment parameters. Specifically, the CAT02 chromatic adaptation transformation matrix is used to calculate the RGB channel drive ratios for the backlight LED based on the color gamut expansion ratio (e.g., to 85% of DCI-P3) in the image display optimization target parameters, combined with the ambient light color temperature data (e.g., 5000K) and the target white point (6500K for the D65 standard illuminant). For example, if adjustment is required from 5000K to 6500K, the blue channel current is increased by 15%, the red channel is reduced by 8%, and the green channel is fine-tuned by 3% to determine the color temperature compensation requirement. Simultaneously, the semantic segmentation results from the key content features are combined to dynamically limit the color temperature adjustment range based on the facial area ratio (e.g., when the face ratio exceeds 30%, the color temperature offset is limited to no more than ±100K to prevent skin color distortion), resulting in the final screen color temperature adjustment value. The optimal global brightness adjustment value is calculated based on the brightness stratification threshold (e.g., 50 nits in the dark area) in the optimization target parameters and the ambient illuminance data (e.g., 2000 lux). The target screen brightness value is calculated using the Stevens power law (perceptual brightness model: S=kIn, n=0.33) For example, when the ambient light is 2000 lux, the target brightness is set to 800 nits to meet the just noticeable difference (JND). The target brightness is dynamically truncated (e.g., not exceeding 950 nits to retain a 10% margin) based on the display's maximum brightness (e.g., 1000 nits) and backlight uniformity parameters. The global image contrast mean is calculated based on the key content features, and the screen brightness target value is fine-tuned to obtain the optimal global brightness adjustment value. The local adjustment parameters are calculated based on the local sharpening strength in the optimization target parameters, combined with the semantic segmentation mask, color attention distribution, and contrast attention distribution in the key content features, to assign different bilateral filter spatial domain parameters (including Gaussian kernel radius and color similarity) to each pixel. In this way, accurate mapping from optimization goals to hardware control instructions can be achieved, providing a direct driving force for the final pixel-level image processing.

[0058] In the above-mentioned method for optimizing image display quality of display equipment, the step S7 performs pixel-level processing on the original image frame data to be displayed based on the display quality adjustment parameters to obtain ambient light adaptive optimized image frame data. Specifically, first, the screen backlight and display driver module are initialized using the screen color temperature adjustment value and the optimal global brightness adjustment value. Then, the image is finely processed by local adjustment parameters, including smoothing the image in the spatial domain and color domain using a bilateral filter to enhance the sharpness and contrast of the local area of the image while retaining edge details. Finally, the processed image data is written to the display buffer, and the image frame data after ambient light adaptive optimization is presented through the display. In this way, the best image display effect can be provided under different ambient light conditions while ensuring the fidelity and visual comfort of the image content.

[0059] In summary, a method for optimizing image display quality for display equipment based on an embodiment of the present application is illustrated, which performs a multi-dimensional joint perceptual analysis of ambient light data and inherent characteristic parameters of the display to quantify the visual sensitivity of the human eye and the performance constraints of the device, and generates a perceptual quality degradation index that characterizes the degree of display quality attenuation. Then, the perceptual quality degradation index and the key content features of the original image frame data to be displayed are further combined to generate image display optimization target parameters that take into account both visual comfort and content expressiveness, ensuring that the image focus is highlighted while compensating for perceptual losses, and on this basis, the screen color temperature adjustment value, the optimal global brightness adjustment value and the local adjustment parameters are calculated to perform pixel-level display optimization processing on the original image frame data. This method can effectively improve the image display quality under different ambient light conditions and different display conditions, providing users with a better visual experience.

[0060] The basic principles of the present invention have been described above in conjunction with specific embodiments. However, it should be noted that the advantages, strengths, and effects mentioned in the present invention are merely illustrative and non-limiting, and should not be construed as necessarily possessed by each embodiment of the present invention. Furthermore, the specific details of the above embodiments are provided for illustrative purposes and to facilitate understanding, and are not intended to be limiting. These details do not necessarily limit the present invention to being implemented using these specific details.

[0061] In the above embodiments, the description of each embodiment has its own emphasis. For parts that are not described or recorded in detail in a certain embodiment, please refer to the relevant description of other embodiments. In the several embodiments provided by the present invention, it should be understood that the disclosed system and method can be implemented in other ways. For example, the system embodiment described above is only schematic. For example, the unit division is only a logical function division, and there may be other division methods in actual implementation. The units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, they may be located in one place, or they may be distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the scheme of this embodiment.

[0062] It will be apparent to those skilled in the art that the present invention is not limited to the details of the exemplary embodiments described above and that the invention can be embodied in other specific forms without departing from the spirit or essential characteristics of the invention. Therefore, the embodiments should be considered in all respects as illustrative and non-restrictive, and the scope of the invention is defined by the appended claims, not the foregoing description, and all variations within the meaning and range of equivalents of the claims are intended to be encompassed therein. Any reference to a figure in a claim should not be construed as limiting the claim to which it relates.

[0063] In addition, it is obvious that the word "comprising" does not exclude other units or steps, and the singular does not exclude the plural. Multiple units stated in the system claims can also be implemented by one unit through software or hardware.

[0064] Finally, it should be noted that the above description has been provided for purposes of illustration and description. Furthermore, the above embodiments are intended only to illustrate the technical solutions of the present invention and are not intended to be limiting. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art will appreciate that the technical solutions of the present invention may be modified or replaced with equivalents without departing from the spirit and scope of the technical solutions of the present invention.

Claims

1. A method for optimizing image display quality of display equipment, characterized in that: include: Acquire ambient light data collected by the ambient light sensor; Obtaining original image frame data to be displayed and inherent characteristic parameters of the display; Extracting key content features from the original image frame data to be displayed; Inputting the ambient light data and the intrinsic characteristic parameters of the display into a perception state evaluation engine to obtain a perception quality degradation indicator; generating image display optimization target parameters based on the perceived quality degradation indicator and the key content features; Calculating a screen color temperature adjustment value, an optimal global brightness adjustment value, and a local adjustment parameter as display quality adjustment parameters based on the image display optimization target parameter and the key content feature; The original image frame data to be displayed is subjected to pixel-level processing based on the display quality adjustment parameter to obtain ambient light adaptive optimized image frame data.

2. The method for optimizing image display quality for display equipment according to claim 1, wherein: Extracting key content features from the original image frame data to be displayed includes: Performing semantic segmentation on the original image frame data to be displayed to obtain a semantic mask image of the original image to be displayed; Performing CIELAB color space conversion and local neighborhood color difference calculation on the original image frame data to be displayed to generate a color attention map of the original image to be displayed; Performing Luma channel conversion and local neighborhood contrast calculation on the original image frame data to be displayed to generate a contrast attention map of the original image to be displayed; The semantic mask map of the original image to be displayed, the color attention map of the original image to be displayed, and the contrast attention map of the original image to be displayed are cascaded along the channel dimension to obtain the key content features.

3. The method for optimizing image display quality for display equipment according to claim 1, wherein: Inputting the ambient light data and the intrinsic characteristic parameters of the display into a perception state evaluation engine to obtain a perception quality degradation indicator, comprising: Performing structured embedded coding on the ambient light data and the display intrinsic characteristic parameter respectively to obtain an ambient light state structured embedded coding vector and a display intrinsic characteristic parameter structured embedded coding vector; Performing multi-scale progressive interactive perception on the ambient light state structured embedded coding vector and the display intrinsic characteristic parameter structured embedded coding vector to obtain an ambient light-display parameter multi-scale progressive interactive perception coding vector; Feature decoding is performed on the ambient light-display parameter multi-scale progressive interaction perceptual coding vector to obtain the perceptual quality degradation indicator.

4. The method for optimizing image display quality for display equipment according to claim 3, wherein: Performing structured embedded coding on the ambient light data to obtain an ambient light state structured embedded coding vector includes: The ambient light data is normalized and then embedded in a fully connected layer to obtain a structured embedded coding vector of the ambient light state.

5. The method for optimizing image display quality for display equipment according to claim 4, wherein: Performing multi-scale progressive interactive perception on the ambient light state structured embedded coding vector and the display intrinsic characteristic parameter structured embedded coding vector to obtain an ambient light-display parameter multi-scale progressive interactive perception coding vector, including: Performing low-level feature interaction coding based on a multi-layer perceptron model on the ambient light state structured embedded coding vector and the display intrinsic characteristic parameter structured embedded coding vector to obtain an ambient light-display parameter low-level joint perceptual coding vector; Performing multi-level high-order feature interaction on the ambient light state structured embedded coding vector and the display intrinsic characteristic parameter structured embedded coding vector to obtain an ambient light-display parameter middle-level joint perception coding vector and an ambient light-display parameter deep-level joint perception coding vector; Multi-scale complementary progressive joint perception is performed on the low-level joint perception coding vector of the ambient light-display parameters, the middle-level joint perception coding vector of the ambient light-display parameters, and the deep-level joint perception coding vector of the ambient light-display parameters to obtain the multi-scale progressive interactive perception coding vector of the ambient light-display parameters.

6. The method for optimizing image display quality for display equipment according to claim 5, characterized in that: Performing multi-level high-order feature interaction on the ambient light state structured embedded coding vector and the display intrinsic characteristic parameter structured embedded coding vector to obtain an ambient light-display parameter middle-level joint perception coding vector and an ambient light-display parameter deep-level joint perception coding vector, including: Performing multi-level implicit feature extraction on the ambient light state structured embedded coding vector and the display intrinsic characteristic parameter structured embedded coding vector to obtain an ambient light state feature middle-level implicit coding vector, a display intrinsic characteristic feature middle-level implicit coding vector, an ambient light state feature deep-level implicit coding vector, and a display intrinsic characteristic feature deep-level implicit coding vector; Performing a mid-level feature interaction based on a cross-attention mechanism on the mid-level implicit coding vector of the ambient light state feature and the mid-level implicit coding vector of the display intrinsic characteristic feature to obtain the mid-level joint perceptual coding vector of the ambient light-display parameter; A deep-level feature interaction based on linear projection gated interaction is performed on the deep-level implicit coding vector of the ambient light state feature and the deep-level implicit coding vector of the display intrinsic characteristic feature to obtain the deep-level joint perception coding vector of the ambient light-display parameter.

7. The method for optimizing image display quality for display equipment according to claim 6, wherein: The method performs multi-scale complementary progressive joint perception on the low-level joint perception coding vector of the ambient light-display parameter, the middle-level joint perception coding vector of the ambient light-display parameter, and the deep-level joint perception coding vector of the ambient light-display parameter to obtain the multi-scale progressive interactive perception coding vector of the ambient light-display parameter, including: Performing complementary interactive fusion based on a gating mechanism on the low-level joint perception coding vector of the ambient light-display parameter and the medium-level joint perception coding vector of the ambient light-display parameter to obtain a low-level joint perception coding vector of the ambient light-display parameter; A cross-level interaction based on a cross-attention mechanism is performed on the low-level joint perception coding vector of the ambient light-display parameter and the deep-level joint perception coding vector of the ambient light-display parameter to obtain the multi-scale progressive interaction perception coding vector of the ambient light-display parameter.

8. The method for optimizing image display quality for display equipment according to claim 7, wherein: Performing cross-level interaction based on a cross-attention mechanism on the low-level joint perception coding vector of the ambient light-display parameter and the deep-level joint perception coding vector of the ambient light-display parameter to obtain the multi-scale progressive interaction perception coding vector of the ambient light-display parameter, including: Obtaining a query embedding matrix, a key embedding matrix, and a value embedding matrix, and performing weight distribution dynamic optimization based on multi-objective collaborative constraints on the query embedding matrix, the key embedding matrix, and the value embedding matrix to obtain an optimized query embedding matrix, an optimized key embedding matrix, and an optimized value embedding matrix; Embedding the low-level joint perceptual coding vector of the ambient light-display parameters using the optimized query embedding matrix to obtain a query vector, and embedding the deep-level joint perceptual coding vector of the ambient light-display parameters using the optimized key embedding matrix and the optimized value embedding matrix to obtain a key vector and a value vector respectively; A cross-level interaction based on a converter architecture is performed on the query vector, the key vector, and the value vector to obtain the ambient light-display parameter multi-scale progressive interaction perception encoding vector.

9. The method for optimizing image display quality for display equipment according to claim 8, characterized in that: Performing feature decoding on the multi-scale progressive interaction perceptual coding vector of the ambient light-display parameter to obtain the perceptual quality degradation indicator includes: The ambient light-display parameter multi-scale progressive interaction perceptual coding vector is input into a feature decoder based on a multi-layer perceptron to obtain the perceptual quality degradation indicator.

Images