Three-dimensional display equipment color gamut mapping method based on multi-dimensional feature fusion and intelligent optimization

Abstract

The invention discloses a three-dimensional display equipment color gamut mapping method based on multi-dimensional feature fusion. A five-dimensional color space model fusing brightness, saturation, hue, depth and visual angle parameters is constructed. On the basis, a neural network architecture of a deep fusion learning and optimization algorithm is applied, so that the method has the capabilities of dynamically learning and optimizing a color gamut mapping function, and adaptive adjustment of display content and dynamic change of a visual angle of a user is realized. The method further comprises a flexible configuration file management system, the system greatly improves the operation flexibility and the use convenience of color gamut mapping through a storage and calling mechanism, and the efficiency of color management is remarkably improved in diversified three-dimensional display equipment and application scenes; finally, the invention provides a high-precision and self-adaptive color gamut mapping scheme suitable for the three-dimensional display equipment, and the remarkable development of the three-dimensional display technology in the fields of color rendition and visual experience is effectively promoted.

Description

Technical Field

[0001] The present invention belongs to the field of three-dimensional display technology, and in particular relates to a color gamut mapping method for a three-dimensional display device based on multi-dimensional feature fusion and intelligent optimization. Background Art

[0002] Current three-dimensional display devices, due to their different screen materials, light-emitting mechanisms, color gamut coverage, and viewing angle and depth perception characteristics, generally encounter color distortion and low mapping efficiency when performing the conversion from the original color gamut to the target color gamut. The uniqueness of three-dimensional display technology lies in its ability to provide immersive depth information and multi-angle visual experience, which undoubtedly increases the complexity of color management. For example, true three-dimensional stereoscopic display technology that does not require additional auxiliary equipment can support high-resolution dynamic three-dimensional images and moving parallax effects. However, the current color gamut mapping technology mostly uses a global conversion function, which is difficult to cope with the differentiated requirements of different devices in color characteristics, especially when dealing with complex changes in depth and viewing angle dimensions. In addition, the color gamut mapping process often requires a lot of computing resources, resulting in slow mapping speed.

[0003] Specifically, the limitations of current methods are mainly reflected in the following aspects. Although color lookup table (LUT) mapping can quickly retrieve color values, its accuracy is limited by the size of the lookup table, and color distortion is prone to occur in scenes with significant depth and viewing angle changes. The matrix color conversion method is based on linear assumptions and has difficulty matching the nonlinear characteristics of the actual color space, and cannot accurately present the complex color characteristics of 3D display devices. Although the color management module (CMM) provides complex algorithms, the effect and efficiency vary from module to module, and it is not suitable for real-time processing of dynamic 3D displays. Model-based color conversion methods rely on accurate measurement data and appropriate models, but most existing models ignore the impact of depth and viewing angle on color, resulting in a large deviation between the conversion results and actual visual perception. Dynamic color gamut mapping is adjustable, but the computation is complex and demanding, making it difficult to adapt to real-time adjustment of multi-dimensional color spaces. Hardware acceleration improves efficiency, but it relies on specific hardware and is expensive, limiting its widespread application. Global conversion functions do not fully consider the color gamut characteristics of different devices, especially the differences in depth and viewing angle, making it difficult to achieve ideal consistency in color conversion results. Summary of the Invention

[0004] In view of this, the purpose of the present invention is to provide a color gamut mapping method for three-dimensional display devices based on multi-dimensional feature fusion and intelligent optimization. Through high-precision spectral acquisition, multi-dimensional feature fusion, intelligent optimization algorithm and dynamic adjustment mechanism, high-precision and adaptive color gamut mapping is achieved, which significantly improves color management efficiency and visual experience.

[0005] A method for color gamut mapping of a three-dimensional display device based on multi-dimensional feature fusion and intelligent optimization includes the following steps:

[0006] Collect spectral energy distribution data of 3D display devices under multiple viewing angles, different parallax depths and different brightness conditions;

[0007] According to the collected spectral energy distribution data, the corresponding XYZ chromaticity value is calculated using the color matching function to establish a fusion lightness L * , Saturation C * ab , Hue H, Viewing Angle and a five-dimensional color space model of disparity depth d;

[0008] The five-dimensional color space data of the five-dimensional color space model are used as input features of the neural network. The output layer of the neural network outputs the target color gamut coordinates, which means that the output layer generates the color coordinates of the target color gamut corresponding to the input. These coordinates indicate the position of the color in the target color gamut to dynamically learn and optimize the color gamut mapping function.

[0009] According to the optimized color gamut mapping function, the mapping between the original color gamut of the display device and different color gamuts is realized.

[0010] Preferably, the corresponding XYZ chromaticity values ​​are calculated based on the collected spectral energy distribution data using a color matching function, wherein the color matching function uses CIE 1931 2°C MFs, CIE 1964 10°C MFs, and CIE 2006 (1°-10° / 20-70 years) CMFs; the calculated XYZ chromaticity values ​​of red, green, blue, and gray are converted into xy color coordinates, and the XYZ chromaticity values ​​are mapped to the CIE xy color space using a chromaticity coordinate formula; in the CIE xy color space, the original color gamut range of the display device and the corresponding white point coordinates of the display device are determined using the xy color coordinates.

[0011] Preferably, in establishing a five-dimensional color space model, the lightness L * Extracted directly from the CIELAB color space; saturation C * ab By calculating a * and b * The Euclidean distance of the components is obtained, and the hue H is calculated by the inverse tangent function. The result is normalized to the range of [0,360) degrees to represent the hue angle, specifically including:

[0012] For each spectral data point collected by the spectrometer, the corresponding X, Y, and Z values ​​are calculated by integrating and multiplying by the CIE 1931 2°C MFs:

[0013]

[0014] Among them, X, Y, and Z represent the tristimulus values ​​of the light source or object, which are used to quantitatively describe the color characteristics of the light source or object within the visible spectrum; Φ(λ) is the color stimulus function, S(λ) represents the relative spectral power distribution of the light source, which describes the light energy output of the light source at different wavelengths; x(λ), y(λ), and z(λ) represent the spectral response capabilities of the human eye's cone cells to the three types of cone cells that are sensitive to red, green, and blue, respectively; k represents the normalization constant; after calculating the tristimulus values, they are further converted into CIExy chromaticity coordinates:

[0015]

[0016] Wherein, the x and y values ​​are defined according to the chromaticity coordinates in the CIE chromaticity diagram, and the z value is calculated from the x and y values, i.e., x+y+z=1;

[0017] Calculate L in CIELAB color space * 、a * and b * value:

[0018]

[0019] Where, F(X / X n )、F(Y / Y n )、F(Z / Z n ) is composed of piecewise functions, and the calculation process is shown in formula (5):

[0020]

[0021] Where X, Y, and Z represent the three stimulus values ​​of color stimulation. X n 、 Y n 、 Z n It represents the tristimulus value when the CIE standard illuminant illuminates the surface of a completely diffuse reflecting object, that is, the color of the lighting source; then formula (4) is converted into a common form, as shown in formula (6):

[0022]

[0023] Saturation The calculation formula of hue H is shown in formula (7):

[0024]

[0025] Preferably, in establishing a five-dimensional color space model, the viewing angle The determination of is achieved by analyzing the observer's line of sight, where θ represents the horizontal azimuth angle of the observer relative to the display screen, and φ represents the vertical elevation angle. The horizontal azimuth angle θ is obtained by calculating the angle between the observer's line of sight projected on the XY plane and the positive X axis, using the inverse tangent function atan2(Y,X) to cover all quadrants. The vertical elevation angle φ is determined by calculating the angle between the line of sight and the XY plane, using Function, which can accurately reflect the degree of the observer's line of sight looking up or down relative to the horizontal plane. The calculation formula is shown in formula (8):

[0026]

[0027] Where X represents the horizontal offset of the observer's eye relative to the center of the display screen in the horizontal direction, Y represents the vertical offset of the observer's eye relative to the center of the display screen in the vertical direction, and Z represents the distance of the observer's eye relative to the center of the display screen in the depth direction. It is the projection length of the observer's line of sight on the XY plane.

[0028] Preferably, in establishing the five-dimensional color space model, the parallax depth influence factor f d (d), which is calculated as f d (d) = 1 + k*e (-λd) , where: the coefficient k is the depth influence coefficient, which is used to adjust the sensitivity of the effect of parallax depth on color intensity, the parameter λ is the depth attenuation coefficient, which determines the rate at which color intensity decays with increasing parallax depth, and the variable d represents the parallax depth coordinate of the color sample, which quantifies the perceived depth of the color in three-dimensional space.

[0029] Preferably, after the initial training of the neural network model is completed, the model is verified, and the model performance is preliminarily evaluated by comparing the difference between the predicted color and the actual color through the color difference formula, and the model parameters are adjusted according to the verification results.

[0030] Preferably, the deep neural network adopts a convolutional neural network CNN or a recurrent neural network RNN ​​to process complex nonlinear relationships in color gamut mapping.

[0031] Preferably, the color difference formula adopts CIELAB or CIEDE2000 color difference formula.

[0032] Preferably, a configuration file management system is implemented to store, call and dynamically update the color gamut mapping parameters, specifically including: designing a parameterized management system that can store, retrieve and update the color gamut mapping parameters in a parameterized manner to ensure flexible configuration and efficient management of the parameters; determining the data storage format of the configuration file to ensure that the color gamut mapping parameters can be stored in a structured and accessible manner, and using an optimized data storage format to achieve fast access and low-latency update of the color gamut mapping parameters; using an adaptive adjustment algorithm to achieve real-time dynamic refresh of the parameters, which can automatically adjust the color gamut mapping parameters according to display content and environmental changes to maintain optimal color performance.

[0033] Preferably, the optimized color gamut mapping technology is integrated into the 3D display device, and a comprehensive performance verification test is conducted to evaluate its performance in different devices and application scenarios, including: adjusting the color gamut mapping parameters according to the characteristics of different 3D display devices to ensure the consistency and accuracy of color reproduction, so that the color gamut mapping technology can work effectively on a variety of devices, and conducting a detailed evaluation of the conversion effect from the original color gamut to the target color gamut to determine the accuracy and color fidelity of the mapping process.

[0034] The present invention has the following beneficial effects:

[0035] The method proposed in this paper achieves high-precision, adaptive mapping of the color gamut of three-dimensional display devices, significantly improving color management efficiency and visual experience. Through multidimensional feature fusion and intelligent optimization algorithms, this method ensures consistent and accurate color reproduction across varying viewing angles, parallax depths, and brightness conditions, providing users with a more natural and realistic visual experience. Furthermore, the implementation of a profile management system enhances the flexibility and convenience of color gamut mapping technology, enabling its widespread application across a wide range of display devices and scenarios, demonstrating significant practical value and market potential. BRIEF DESCRIPTION OF THE DRAWINGS

[0036] Figure 1 A schematic flow chart of a method for color gamut mapping of a three-dimensional display device based on multi-dimensional feature fusion and intelligent optimization provided in Example 1 of the present invention;

[0037] Figure 2 This is a schematic diagram of multi-angle measurement according to Example 1 of the present invention;

[0038] Figure 3 The LCH color space described in Example 1 of the present invention;

[0039] Figure 4 The horizontal and vertical viewing angles of the observer described in Example 1 of the present invention;

[0040] Figure 5This is the neural network model construction process described in Example 1 of the present invention;

[0041] Figure 6 The standard color gamut space described in Example 1 of the present invention;

[0042] Figure 7 This is the application process on different three-dimensional display devices described in Example 2 of the present invention. DETAILED DESCRIPTION

[0043] The present invention is described in detail below with reference to the accompanying drawings and embodiments.

[0044] The purpose of the present invention is to provide a color gamut mapping method for a three-dimensional display device based on multi-dimensional feature fusion and intelligent optimization.

[0045] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the present invention is further described in detail below with reference to the accompanying drawings and specific embodiments. Unless otherwise specified, the equipment used in the following embodiments are all conventional equipment available on the market in this field.

[0046] Example 1:

[0047] like Figure 1 As shown, a color gamut mapping method for a three-dimensional display device based on multi-dimensional feature fusion and intelligent optimization in this embodiment includes the following steps. Figure 2 Measure the diagram for a 3D display device. Figure 3 For the neural network model building process, Figure 4 It is a standard color gamut space.

[0048] Step 101: collecting spectral energy distribution data of a three-dimensional display device under multiple viewing angles, parallax depths, and brightness conditions;

[0049] Step 102: Calculate the corresponding XYZ chromaticity value using the color matching function according to the spectral energy distribution data collected in step (1) to establish a fusion lightness (L * ), saturation (C * ab ), hue (H), observer's horizontal and vertical viewing angles (θ, φ) and parallax depth (d);

[0050] Step 103: Design a neural network model that deeply integrates learning and optimization algorithms, and use the collected data set for training to dynamically learn and optimize the color gamut mapping function;

[0051] Step 104: During the model training process, the color gamut mapping function is optimized to ensure that it adapts to real-time changes in display content and user viewing angles;

[0052] Step 105: Implement a profile management system to store, call, and dynamically update gamut mapping parameters.

[0053] Step 106: Integration and performance verification testing: Integrate the optimized color gamut mapping technology into the 3D display device and conduct comprehensive performance verification testing to evaluate its performance in different devices and application scenarios, and achieve mapping between the original color gamut of the display device and standard color gamut spaces such as sRGB, DCI-P3, and BT.2020.

[0054] Wherein, step 101 specifically includes:

[0055] Adjust the 3D display device to the best working state, set the scanning parameters, and perform multi-angle spectrum measurement with an interval of 10° within the range of 0° to 180°. The measurement diagram is shown in the figure below. Figure 2 shown.

[0056] Determines the parallax depth from 0mm to 100mm in 10mm intervals to simulate different display depths.

[0057] Set the brightness level from 10% to 100% in 10% intervals to simulate different ambient brightness conditions.

[0058] At each parallax depth and brightness level combination, a spectrum analyzer was used to record the spectral power distribution data of red, green, blue, and white light.

[0059] Display devices include but are not limited to 3D displays, virtual reality (VR) devices, augmented reality (AR) devices, etc.

[0060] Wherein, step 102 specifically includes:

[0061] Calculate the measured XYZ chromaticity values ​​of red, green, blue, and gray using a color matching function based on the red, green, blue, and white spectral energy distribution data collected in step 101, and determine the original color gamut of the display device and its corresponding white point;

[0062] The color matching functions include CIE 1931 2°, CIE 1964 10°, and CIE 2006 (1°-10° / 20-70 years); the original color gamut and its corresponding white point are XYZ chromaticity values ​​of red, green, blue and gray, and the corresponding xy color coordinates are calculated using the chromaticity coordinate formula, and the color range is determined by constructing in the CIE xy color space.

[0063] Calculate the lightness L in the CIELAB color space * , chromaticity coordinate a * and b * , and then get the saturation C *ab and hue H, LCH color space such as Figure 3 As shown;

[0064] Lightness L * Extracted directly from the CIELAB color space, saturation C * ab By calculating a * and b * The Euclidean distance of the components is obtained, and the hue H is calculated by the inverse tangent function atan2. The result is usually normalized to [0,360) degrees to represent the hue angle;

[0065] The viewing angle (θ, φ) is determined by analyzing the observer's line of sight, where θ represents the horizontal azimuth angle of the observer relative to the display screen, and φ represents the vertical elevation angle. The horizontal viewing angle and vertical viewing angle of the observer are as follows: Figure 4 As shown;

[0066] For the parallax depth factor f d (d), which is calculated as f d (d) = 1 + k*e -λd , which quantifies the perceived depth of color in three-dimensional space.

[0067] This process can be expressed as:

[0068] For each spectral data point collected by the spectrometer, the corresponding X, Y, and Z values ​​are calculated by integrating and multiplying by the CIE 1931 2°C MFs, as shown in formula (1). This process can be expressed as:

[0069]

[0070] Among them, X, Y, and Z represent the tristimulus values ​​of the light source or object, which are used to quantitatively describe the color characteristics of the light source or object within the visible spectrum. Ф(λ) is the color stimulus function (which quantifies the response intensity of the human eye to light at different wavelengths and is used to simulate the perception process under different color light stimuli), and S(λ) represents the relative spectral power distribution of the light source, which describes the light energy output of the light source at different wavelengths. x(λ), y(λ), and z(λ) respectively characterize the spectral response capabilities of the human eye's cone cells to the three types of cone cells that are sensitive to red, green, and blue. k represents the normalization constant, and the calculation formula is shown in formula (2). Its function is to standardize the Y value of the tristimulus values ​​of the lighting source to 100 to ensure that the brightness factor of the light source is 100, thereby maintaining a consistent comparison standard in colorimetric analysis. After calculating the tristimulus values, they are further converted into CIE xy chromaticity coordinates. The calculation formula is shown in formula (3):

[0071]

[0072] The x and y values ​​are defined according to the chromaticity coordinates in the CIE chromaticity diagram, and the z value can be calculated from the x and y values, that is, x+y+z=1.

[0073] Through the above integration, we get the XYZ chromaticity value of each data point, which will be used to calculate the L in CIELAB color space. * 、a * and b * value, and then get the saturation C a * b And hue H, the calculation formula is shown in formula (4):

[0074]

[0075] Where, F(X / X n )、F(Y / Y n )、F(Z / Z n ) is composed of piecewise functions, and the calculation process is shown in formula (5):

[0076]

[0077] In the formula, X, Y, and Z represent the three stimulus values ​​of color stimulation, X n 、Y n 、Z n Indicates the tristimulus values ​​when the CIE standard illuminant illuminates the surface of a completely diffuse reflecting object, that is, the color of the lighting source. In most cases, the X, Y, and Z tristimulus values ​​are all greater than 1, then calculate L * 、a * and b * The value is converted into a common form, as shown in formula (6):

[0078]

[0079] Where, L * The axis is represented by the white-black axis, which represents the brightness. All colors on the axis are non-colors, the top is white, the bottom is black, and the middle is gray with gradually changing shades; a * The axis is represented as red-green axis, + a * is (magenta), - a * is green; b * The axis is represented as yellow-blue axis, + b * Yellow, - b * is blue. Saturation The calculation formula of hue H is shown in formula (7):

[0080]

[0081] In a three-dimensional coordinate system with the center of the display screen as the origin, the coordinates of the observer's eye are (X, Y, Z). The viewing angle (θ, φ) is determined by analyzing the observer's line of sight, where θ represents the horizontal azimuth of the observer relative to the display screen, and φ represents the vertical elevation angle. The horizontal azimuth angle θ is obtained by calculating the angle between the observer's line of sight projected onto the XY plane and the positive X axis, using the inverse tangent function atan2(Y, X) to cover all quadrants. The vertical elevation angle φ is determined by calculating the angle between the line of sight and the XY plane, using atan2(Z, sqrt(X 2 +Y 2 )) function, which can accurately reflect the degree of the observer's line of sight looking up or down relative to the horizontal plane. The calculation formula is shown in formula (8):

[0082]

[0083] Where X represents the horizontal offset of the observer's eyes relative to the center of the display screen in the horizontal direction, Y represents the vertical offset of the observer's eyes relative to the center of the display screen in the vertical direction, and Z represents the distance of the observer's eyes relative to the center of the display screen in the depth direction. sqrt(X 2 +Y 2 ) is the projection length of the observer's line of sight on the XY plane.

[0084] For the parallax depth factor f d (d), the calculation formula is shown in formula (9):

[0085] f d (d) = 1 + k*e (-λd) (9)

[0086] Among them, the coefficient k is the depth influence coefficient, which is used to adjust the sensitivity of the influence of parallax depth on color intensity. The parameter λ is the depth attenuation coefficient, which determines the rate at which color intensity decays with increasing parallax depth. The variable d represents the parallax depth coordinate of the color sample, which quantifies the perceived depth of color in three-dimensional space, thus providing a comprehensive and accurate framework for color gamut mapping of three-dimensional display devices.

[0087] Wherein, step 103 specifically includes:

[0088] Build a multilayer perceptron (MLP) neural network model, which is a feedforward neural network consisting of an input layer, one or more hidden layers, and an output layer. This architecture is suitable for learning complex nonlinear relationships in color gamut mapping.

[0089] The input layer receives five-dimensional color space coordinates, which means that the input layer can accept five-dimensional color space data As input features of the neural network;

[0090] The hidden layer uses the ReLU (Rectified Linear Unit) activation function, which effectively introduces nonlinear factors while maintaining computational efficiency. To reduce overfitting, a Dropout layer is added to the hidden layer. This layer randomly "drops" the output of some neurons during training, forcing the network to learn more robust feature representations.

[0091] The output layer outputs the target color gamut coordinates, which means that the output layer generates the color coordinates of the target color gamut corresponding to the input. These coordinates indicate the position of the color in the target color gamut.

[0092] In the process of building the neural network model, we considered the fusion strategy of multiple deep learning algorithms, including convolutional neural networks (CNNs) and recurrent neural networks (RNNs). These algorithms are widely recognized for handling the complex nonlinear relationships in color gamut mapping. However, given the structured nature of the input data, we ultimately adopted a multilayer perceptron (MLP) model, as it is more suitable for processing this type of data structure and can effectively extract and transform color features.

[0093] To optimize the weights and biases of the neural network model, the Adam optimizer is used, which is an adaptive moment estimation optimization algorithm that combines the advantages of traditional gradient descent and momentum methods;

[0094] The collected data is preprocessed, including data normalization, denoising, and enhancement, to improve the effectiveness of neural network training and the generalization ability of the model. The neural network model is initially trained using the preprocessed data set to enable the model to learn how to map the five-dimensional color space coordinates to the target color gamut;

[0095] During the training process, a dynamic learning mechanism is introduced to lay the foundation for the real-time optimization in step (4). After the initial training is completed, the model verification process is performed to compare the difference between the color values ​​predicted by the model and the actual color values ​​through the color difference formula, and the performance of the model is preliminarily evaluated. Based on the verification results, the model parameters are fine-tuned, including but not limited to the number of hidden layer neurons, Dropout ratio, and learning rate, in order to further improve the accuracy and computational efficiency of the model; the color difference formula includes but is not limited to CIELAB and CIEDE2000 color difference formulas;

[0096] Wherein, step 104 specifically includes:

[0097] During model training, a real-time feedback mechanism is established that monitors changes in display content and user perspective and transmits these changes as input signals to the neural network model.

[0098] Dynamically adjust the weights and biases of the neural network model based on the input signal to optimize the color gamut mapping function in real time to ensure the accuracy of color reproduction;

[0099] A continuous learning strategy is implemented for different display content, including video games, movies, text, etc., and user perspectives, including frontal viewing and sideways viewing, so that the neural network model can continue to learn and optimize in the ever-changing display environment and user behavior without having to retrain the entire model. The method includes taking real-time collected user perspective change data and display content features as input and feeding them back to the neural network model in real time; using small batch learning technology to incrementally train the model, that is, during the user's viewing process, the model learns new color mapping patterns in the form of small batches of data; and using sliding window technology to maintain the model's learning memory of recent data while gradually forgetting old data to adapt to long-term trends and short-term changes in color mapping.

[0100] During model execution, the performance of the color gamut mapping function is monitored in real time, and the difference between the predicted color and the actual color is quantified using the color difference formula. The model parameters are iteratively adjusted to minimize the loss function, and the learning rate is dynamically adjusted using a learning rate scheduling method to improve training efficiency and model performance.

[0101] Wherein, step 105 specifically includes:

[0102] Design a parameterized management system that can store, retrieve, and update color gamut mapping parameters in a parameterized manner to ensure flexible configuration and efficient management of parameters;

[0103] Determine the data storage format for the configuration file, ensuring that the gamut mapping parameters can be stored in a structured and accessible manner, such as XML, JSON, or binary format;

[0104] Use optimized data storage formats, such as hash tables or tree structures, to enable fast access and low-latency updates of gamut mapping parameters;

[0105] Adaptive adjustment algorithms are used to achieve real-time dynamic refresh of parameters, automatically adjusting color gamut mapping parameters according to display content and environmental changes to maintain optimal color performance.

[0106] Wherein, step 106 specifically includes:

[0107] Adjust the color gamut mapping parameters according to the characteristics of different 3D display devices to ensure the consistency and accuracy of color reproduction, so that the color gamut mapping technology can work effectively on a variety of devices;

[0108] Perform detailed evaluation of the conversion from the original color gamut to the target color gamut (e.g., sRGB, DCI-P3, BT.2020) to determine the accuracy and color fidelity of the mapping process.

[0109] Example 2

[0110] like Figure 7 As shown, this embodiment further verifies the application process of the method of the present invention on different three-dimensional display devices, including the following steps.

[0111] Step 201: Data collection and equipment characteristics analysis;

[0112] Step 202: Neural network model construction and training;

[0113] Step 203: Optimizing color gamut mapping parameters;

[0114] Step 204: Real-time feedback and system optimization.

[0115] Wherein, step 201 specifically includes:

[0116] A 3D projector, a VR head-mounted display, and AR smart glasses were selected as experimental equipment. Under a standard test environment, a spectrum analyzer was used to collect spectral energy distribution data for each device at multiple angles, parallax depth, and brightness conditions, ranging from 0° to 180° and at 10° intervals. Key performance indicators for each device were recorded, including display technology, color gamut coverage, resolution, brightness, and contrast.

[0117] Wherein, step 202 specifically includes:

[0118] Referring to step 103 in Example 1, a multi-layer perceptron (MLP) neural network model is constructed, and the input layer receives the five-dimensional color space coordinates (L * , C * ab ,H,θ,φ,f d (d)), the hidden layer contains 128 neurons, uses the ReLU activation function, and the output layer outputs the target color gamut coordinates.

[0119] The neural network model is trained using the collected data set to ensure that the model can learn how to map the five-dimensional color space coordinates to the target color gamut.

[0120] Wherein, step 203 specifically includes:

[0121] According to step 106 of Example 1, the color gamut mapping parameters are adjusted based on the color gamut characteristics of each device to optimize color reproduction. The conversion effect from the original color gamut to the target color gamut is evaluated in detail using the CIELAB or CIEDE2000 color difference formula to ensure color fidelity.

[0122] Wherein, step 204 specifically includes:

[0123] The real-time feedback mechanism of Example 1 is used to dynamically adjust the color gamut mapping parameters to adapt to changes in display content and user viewing angle.

[0124] According to the parameterized management solution of Example 1, the configuration file management system is used to implement rapid access and dynamic updating of color gamut mapping parameters, thereby improving the flexibility and response speed of the system.

[0125] This embodiment verifies the effectiveness and practicality of the color gamut mapping method of the present invention in processing different types of three-dimensional display devices through specific experiments. This method significantly improves color management efficiency and visual experience through multi-dimensional feature fusion and intelligent optimization algorithm, and provides an innovative solution for the development of three-dimensional display technology. Through comprehensive processing of device characteristic analysis, personalized model training, color gamut mapping parameter adjustment, environmental adaptability, user scenario optimization, real-time feedback and optimization, cross-device consistency and hardware calibration, it is ensured that each device can provide the best color performance and user experience. The color reproduction error of all devices is kept within CIEDE2000 <2.0, which proves the efficiency and accuracy of the method of the present invention.

[0126] The display device used in the present invention may include a variety of three-dimensional display devices, including 3D displays, virtual reality (VR) devices, augmented reality (AR) devices, etc.; the original color gamut space division standard can be changed according to the color characteristics of the display device; the color gamut mapping method obtained by the method of the present invention can be applied to color gamut mapping and prediction between three-dimensional display devices with different color rendering principles. In summary, the above is only a preferred embodiment of the present invention and is not intended to limit the scope of protection of the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the scope of protection of the present invention.

Claims

1. A color gamut mapping method for a three-dimensional display device based on multi-dimensional feature fusion and intelligent optimization, characterized in that: The following steps are involved: Collect spectral energy distribution data of 3D display devices under multiple viewing angles, different parallax depths and different brightness conditions; According to the collected spectral energy distribution data, the corresponding XYZ chromaticity value is calculated using the color matching function to establish a fusion lightness L * , Saturation C * ab , Hue H, Viewing Angle and a five-dimensional color space model of disparity depth d; The five-dimensional color space data of the five-dimensional color space model are used as input features of the neural network. The output layer of the neural network outputs the target color gamut coordinates, which means that the output layer generates the color coordinates of the target color gamut corresponding to the input. These coordinates indicate the position of the color in the target color gamut to dynamically learn and optimize the color gamut mapping function. According to the optimized color gamut mapping function, the mapping between the original color gamut of the display device and different color gamuts is realized.

2. The color gamut mapping conversion process method for a three-dimensional display device according to claim 1, characterized in that: Corresponding XYZ chromaticity values ​​are calculated based on the collected spectral energy distribution data using a color matching function, wherein the color matching function adopts CIE 1931 2°C MFs, CIE 1964 10°C MFs, and CIE 2006 (1°-10° / 20-70 years) CMFs. The calculated XYZ chromaticity values ​​of red, green, blue, and gray are converted into xy color coordinates using a chromaticity coordinate formula to map the XYZ chromaticity values ​​to the CIE xy color space. In the CIE xy color space, the original color gamut range of the display device and the corresponding white point coordinates of the display device are determined using the xy color coordinates.

3. The color gamut mapping conversion process method for a three-dimensional display device according to claim 2, characterized in that: In establishing the five-dimensional color space model, the lightness L * Extracted directly from the CIELAB color space; saturation C * ab By calculating a * and b * The Euclidean distance of the components is obtained, and the hue H is calculated by the inverse tangent function. The result is normalized to the range of [0,360) degrees to represent the hue angle, specifically including: For each spectral data point collected by the spectrometer, the corresponding X, Y, and Z values ​​are calculated by integrating and multiplying by the CIE 1931 2°C MFs: Among them, X, Y, and Z represent the tristimulus values ​​of the light source or object, which are used to quantitatively describe the color characteristics of the light source or object within the visible spectrum; Φ(λ) is the color stimulus function, S(λ) represents the relative spectral power distribution of the light source, which describes the light energy output of the light source at different wavelengths; x(λ), y(λ), and z(λ) represent the spectral response capabilities of the human eye's cone cells to the three types of cone cells that are sensitive to red, green, and blue, respectively; k represents the normalization constant. After calculating the tristimulus values, they are further converted into CIE xy chromaticity coordinates: Wherein, the x and y values ​​are defined according to the chromaticity coordinates in the CIE chromaticity diagram, and the z value is calculated from the x and y values, i.e., x+y+z=1; Calculate L in CIELAB color space * 、a * and b * value: Where, F(X / X n )、F(Y / Y n )、F(Z / Z n ) is composed of piecewise functions, and the calculation process is shown in formula (5): Where X, Y, and Z represent the three stimulus values ​​of color stimulation. X n 、 Y n 、 Z n It represents the tristimulus value when the CIE standard illuminant illuminates the surface of a completely diffuse reflecting object, that is, the color of the lighting source; then formula (4) is converted into a common form, as shown in formula (6): Saturation The calculation formula of hue H is shown in formula (7):

4. The color gamut mapping conversion process method for a three-dimensional display device according to claim 4, characterized in that: In building a five-dimensional color space model, the viewing angle The determination of is achieved by analyzing the observer's line of sight, where θ represents the horizontal azimuth angle of the observer relative to the display screen, and φ represents the vertical elevation angle. The horizontal azimuth angle θ is obtained by calculating the angle between the observer's line of sight projected on the XY plane and the positive X axis, using the inverse tangent function atan2(Y,X) to cover all quadrants. The vertical elevation angle φ is determined by calculating the angle between the line of sight and the XY plane, using Function, which can accurately reflect the degree of the observer's line of sight looking up or down relative to the horizontal plane. The calculation formula is shown in formula (8): Where X represents the horizontal offset of the observer's eye relative to the center of the display screen in the horizontal direction, Y represents the vertical offset of the observer's eye relative to the center of the display screen in the vertical direction, and Z represents the distance of the observer's eye relative to the center of the display screen in the depth direction. It is the projection length of the observer's line of sight on the XY plane.

5. The color gamut mapping conversion process method for a three-dimensional display device according to claim 4, characterized in that: In establishing the five-dimensional color space model, the parallax depth factor f d (d), which is calculated as f d (d) = 1 + k*e (-λd) , where: the coefficient k is the depth influence coefficient, which is used to adjust the sensitivity of the effect of parallax depth on color intensity, the parameter λ is the depth attenuation coefficient, which determines the rate at which color intensity decays with increasing parallax depth, and the variable d represents the parallax depth coordinate of the color sample, which quantifies the perceived depth of the color in three-dimensional space.

6. The color gamut mapping conversion process method for a three-dimensional display device according to claim 1, characterized in that: After the initial training of the neural network model is completed, the model is verified. The difference between the predicted color and the actual color is compared through the color difference formula to preliminarily evaluate the model performance, and the model parameters are adjusted according to the verification results.

7. The color gamut mapping conversion process method for a three-dimensional display device according to claim 6, characterized in that: The deep neural network adopts convolutional neural network CNN or recurrent neural network RNN ​​to process the complex nonlinear relationship in color gamut mapping.

8. The color gamut mapping conversion process method for a three-dimensional display device according to claim 7, characterized in that: The color difference formula adopts CIELAB or CIEDE2000 color difference formula.

9. The color gamut mapping conversion process method of a display device according to claim 1, characterized in that: Implement a profile management system to store, call and dynamically update color gamut mapping parameters, specifically including: designing a parameterized management system that can store, retrieve and update color gamut mapping parameters in a parameterized manner to ensure flexible configuration and efficient management of parameters; determining the data storage format of the profile to ensure that the color gamut mapping parameters can be stored in a structured and accessible manner, and using an optimized data storage format to achieve fast access and low-latency update of the color gamut mapping parameters; using an adaptive adjustment algorithm to achieve real-time dynamic refresh of parameters, which can automatically adjust the color gamut mapping parameters according to display content and environmental changes to maintain optimal color performance.

10. The color gamut mapping conversion process method for a three-dimensional display device according to claim 1, characterized in that: The optimized color gamut mapping technology is integrated into 3D display devices, and comprehensive performance verification testing is carried out to evaluate its performance in different devices and application scenarios. Specifically, the color gamut mapping parameters are adjusted according to the characteristics of different 3D display devices to ensure the consistency and accuracy of color reproduction, so that the color gamut mapping technology can work effectively on multiple devices. The conversion effect from the original color gamut to the target color gamut is also evaluated in detail to determine the accuracy and color fidelity of the mapping process.

Images