Method and system for multi-primary display optimization based on individual observer characteristics

Abstract

The application discloses a multi-primary color display optimization method and system based on individual observer characteristics, and belongs to the field of display technology and color management. The application obtains an individual observer variation characteristic matrix composed of an average color matching function and a deviation principal component; then, for a color sample to be reproduced on a multi-primary color display, an optimization target is to minimize the projection error of an output spectrum of the color sample on the multi-primary color display generated by a display drive vector and a projection of a target spectrum of the color sample in a subspace defined by the individual observer variation characteristic matrix, under chroma accurate matching constraints and display drive physical constraints, the display drive vector is optimized and solved to obtain an optimal display drive vector. The application directly optimizes the variation law of the individual observer CMF, and the generated display spectrum can effectively reduce the color perception difference in different eyes and improve the consistency of color reproduction.

Description

Technical Field

[0001] This invention belongs to the field of display technology and color management, and specifically relates to a color reproduction method for multi-primary display (MPD). Background Technology

[0002] With the development of display technology, multi-color displays that go beyond the traditional red, green, and blue (RGB) three primary colors, such as RGBCY, RGBW, or n (n>3) primary color displays using laser light sources, have gradually entered the market. These displays, by increasing the number of primary colors, can present a wider color gamut, thereby reproducing richer and more saturated colors.

[0003] However, multi-color displays also introduce new technical challenges. According to the principles of color science, the triconical cell system of the human eye dictates that any color can be defined by three independent variables (such as CIE XYZ tristimulus values). When the number of primary colors, n, in a display is greater than 3, for a given target XYZ value, there are infinitely many combinations of primary color driving values ​​that can reproduce that color. These different combinations of driving values ​​produce different spectral power distributions (SPDs). Although they appear to be the same color under the CIE 1931 standard observer (i.e., standard metamerism), for individual observers whose color matching function (CMF) differs from the standard observer's, these different spectra will elicit different color perceptions. This phenomenon is known as "observer metamerism."

[0004] In applications requiring extremely high color accuracy, such as medical imaging, professional design, artwork reproduction, and remote color collaboration, observer metamerism can severely impact the reliability and consistency of workflows. Existing technologies typically employ various methods to handle multi-primary-color driven processes, such as:

[0005] 1. Partitioning: The multi-color display is decomposed into multiple three-color displays with different color gamuts. Based on the color coordinates of the input color, a three-color display that covers the current color gamut is found and used for display. However, this method does not consider spectral optimization and may lead to serious metamerism problems for observers.

[0006] 2. Power consumption optimization: Prioritize the use of display primary colors with high luminous efficiency and high visual efficiency of the human eye. This method does not take into account the differences between observers.

[0007] 3. Prioritize specific primary colors: For example, prioritize the use of broad-spectrum primary colors. This method may also lead to serious metamerism problems among observers.

[0008] 4. Spectral matching method: This method attempts to make the spectrum output by the display as close as possible to the original target spectrum. Theoretically, this method can eliminate metamerism, but with a limited number of primary colors in the display, perfect spectral matching is almost impossible to achieve, and its optimization objective (minimizing the root mean square error of the spectrum) does not completely correspond to the error perceived by human vision.

[0009] Therefore, there is an urgent need for a new color reproduction method for multi-color displays that can intelligently select an optimal spectral reproduction scheme while ensuring CIE1931 standard colorimetric matching, so as to minimize the average color perception differences in real people. Summary of the Invention

[0010] This invention aims to solve the observer metamerism problem in the color reproduction of multi-color displays in the prior art, and provides a multi-color display driving method and system that can minimize the average color difference between a series of individual observers and a standard observer.

[0011] The specific technical solution adopted in this invention is as follows:

[0012] In a first aspect, the present invention provides a multi-primary-color display optimization method based on individual observer characteristics, comprising:

[0013] S1. Obtain the individual observer variation feature matrix composed of the average color matching function and the bias principal components; the average color matching function is obtained by averaging the color matching functions of the observer group; the bias principal components are obtained by principal component analysis of the bias functions of the color matching functions of all individual observers in the observer group relative to the average color matching function;

[0014] S2. For a color sample that needs to be reproduced on a multi-color display, the optimization objective is to minimize the projection error between the output spectrum of the color sample generated by the display driving vector on the multi-color display and the target spectrum of the color sample in the subspace defined by the individual observer variation feature matrix. Under the constraints of accurate color matching and display driving physical constraints, the display driving vector is optimized to obtain the optimal display driving vector.

[0015] As a preferred embodiment of the first aspect above, each individual observer in the observer group has the same type of color matching function, selected from one of the LMS color matching function, XYZ color matching function, or RGB color matching function.

[0016] As a preferred embodiment of the first aspect above, both the average color matching function and the bias principal component contain three color channels, and the average color matching function of each color channel and the bias principal component of each color channel are combined in the form of row vectors or column vectors to form the individual observer change feature matrix.

[0017] As a preferred embodiment of the first aspect above, the method for obtaining the biased principal components is as follows:

[0018] For the observer group, the color matching function of each individual observer is subtracted from the average color matching function to obtain the three-channel bias function of each individual observer; principal component analysis is performed on the three-channel bias function of all individual observers in the observer group channel by channel, and a preset number of principal components are extracted for each color channel to be used as bias principal components to construct the individual observer change feature matrix.

[0019] As a preferred embodiment of the first aspect above, in the process of optimizing the display driving vector, for any feasible solution of the display driving vector, the method for calculating the projection error in the objective function is as follows: multiply the primary color spectrum of the multi-primary color display with the display driving vector corresponding to the current feasible solution to obtain the output spectrum of the multi-primary color display, calculate the deviation spectrum between the output spectrum and the target spectrum of the color sample under the target light source, and then use the individual observer change feature matrix as the projection matrix to project the deviation spectrum, and the obtained projection error is used as the objective function value corresponding to the current feasible solution.

[0020] As a preferred embodiment of the first aspect above, the colorimetric precision matching constraint adopts an equality constraint, which is in the form that the tristimulus values ​​obtained by integrating the output spectrum of the color sample and the target spectrum of the color sample through the CIE standard color matching function are equal.

[0021] As a preferred embodiment of the first aspect above, the physical constraint of the display driving is an inequality constraint, which is in the form that each element in the display driving vector of the multi-primary-color display is between 0 and the maximum driving value.

[0022] Secondly, the present invention provides a multi-primary-color display system based on individual observer characteristics, comprising:

[0023] An optimal driving vector storage module is used to store the target spectrum and the optimal display driving vector corresponding to the discrete color node in the three-dimensional color space, as a multi-primary color driving lookup table; the optimal display driving vector is pre-optimized by the discrete color node as the color sample to be reproduced on the multi-primary color display according to the multi-primary color display optimization method based on individual observer characteristics as described in any of the first aspects above.

[0024] The real-time rendering module is used to read the three-channel values ​​of each pixel that needs to be displayed in real time in the three-dimensional color space during the operation of the multi-primary color display, and convert them to obtain the target spectrum of the three-channel values ​​under the target light source. Then, based on the multi-primary color drive lookup table, the optimal display drive vector corresponding to this target spectrum is calculated by interpolation algorithm and sent to the display.

[0025] Thirdly, the present invention provides a computer program product, including a computer program / instruction, which, when executed by a processor, can implement the multi-primary color display optimization method based on individual observer characteristics as described in any of the first aspects above.

[0026] Fourthly, the present invention provides a computer electronic device, which includes a memory and a processor;

[0027] The memory is used to store computer programs;

[0028] The processor is configured to, when executing the computer program, implement the multi-primary-color display optimization method based on individual observer characteristics as described in any of the first aspects above.

[0029] Compared with the prior art, the present invention has the following advantages:

[0030] 1. Significantly reduces interotropic color reproduction: This invention directly optimizes the variation pattern of CMF (Color Filter Mixture) for individual observers, and the generated display spectrum can effectively reduce the differences in color perception among different people and improve the consistency of color reproduction.

[0031] 2. Guarantee standard color accuracy: By using CIE 1931 color matching as a strict equality constraint, this method ensures color accuracy under standard conditions and is fully compatible with existing color workflows.

[0032] 3. Scientific and efficient: Based on a deep understanding of the statistical characteristics of the human visual system, this method transforms the complex perceptual optimization problem into a mature and efficient quadratic programming problem, which is both scientific and practical.

[0033] 4. Flexibility and scalability: The individual observer model can be updated or replaced as needed, and the number of principal components can also be adjusted to suit displays with different numbers of primary colors. Attached Figure Description

[0034] Figure 1 A schematic diagram illustrating the steps of the multi-primary-color display optimization method based on individual observer characteristics provided by the present invention;

[0035] Figure 2 This is the normalized spectrum of the 5-primary-color display in this embodiment of the invention;

[0036] Figure 3 This refers to the LMS color matching function for 100 observers in this embodiment of the invention.

[0037] Figure 4 The average composition and the top 3 principal components of the three channels of LMS in this embodiment of the invention;

[0038] Figure 5 This is the spectrum of a color sample under a D65 light source in an embodiment of the present invention;

[0039] Figure 6 The output spectrum of the 5-primary-color display in this embodiment of the invention;

[0040] Figure 7 This is a flowchart illustrating the calculation of the observer metamerism index in an embodiment of the present invention.

[0041] Figure 8 This is a schematic diagram of the structure of a computer electronic device in an embodiment of the present invention. Detailed Implementation

[0042] To make the above-mentioned objects, features, and advantages of the present invention more apparent and understandable, specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings. Many specific details are set forth in the following description to provide a thorough understanding of the present invention. However, the present invention can be practiced in many other ways different from those described herein, and those skilled in the art can make similar modifications without departing from the spirit of the present invention. Therefore, the present invention is not limited to the specific embodiments disclosed below. Technical features in various embodiments of the present invention can be combined accordingly without mutual conflict.

[0043] like Figure 1 As shown, this invention provides a multi-primary-color display optimization method based on individual observer characteristics, which includes:

[0044] S1. Obtain the individual observer variation feature matrix composed of the average color matching function and the bias principal components; the average color matching function is obtained by averaging the color matching functions of the observer group; the bias principal components are obtained by principal component analysis of the bias functions of the color matching functions of all individual observers in the observer group relative to the average color matching function.

[0045] It should be noted that the observer group referred to in this invention is a group composed of a series of individual observers. Theoretically, the more individual observers it contains, the better it represents the average level. The color matching function (CMF) of each individual observer in the observer group should be of the same type. The three color channels contained therein need to be determined according to the color space corresponding to the color matching function. The color matching function can be selected from LMS color matching function, XYZ color matching function, or RGB color matching function. The color matching function of individual observers has different normalization forms: LMS is the cell's sensitivity function to light wavelength, XYZ is the current international standard color matching function conversion form, and RGB is a color matching function form normalized according to a specific wavelength.

[0046] In addition, the construction of the individual observer change feature matrix in this invention requires the use of the average color matching function and the bias principal component. Both contain three color channels. The average color matching function of each color channel and the bias principal component of each color channel are combined in the form of row vectors or column vectors to form the individual observer change feature matrix, so as to participate in the subsequent projection operation in matrix form.

[0047] It should be noted that, in the embodiments of the present invention, the method for obtaining the above-mentioned deviation principal components is as follows:

[0048] For the observer group, the color matching function of each individual observer is subtracted from the average color matching function to obtain the three-channel bias function of each individual observer; principal component analysis is performed on the three-channel bias function of all individual observers in the observer group channel by channel, and a preset number of principal components are extracted for each color channel to be used as bias principal components to construct the individual observer change feature matrix.

[0049] Additionally, it's important to note that since each individual observer's color matching function and the average color matching function are actually three-channel, the resulting bias function for each individual observer is also three-channel. Assuming there are N individual observers in the observer group, and each individual observer's color matching function (CMF) is denoted as... Let K be the number of discretely sampled wavelength points in the spectrum, then the average color matching function is denoted as... Its calculation formula can be expressed as: .because It actually contains three channels, so the calculation is performed channel by channel. The average color matching function is calculated. Then, it can be used to calculate the three-channel bias function for each individual observer. Its calculation formula is expressed as: .

[0050] Taking RGB three-channel as an example, during principal component analysis (PCA), the bias function of each individual observer in each channel (R, G, and B channels) needs to be analyzed separately, retaining a predetermined number of M principal components (eigenvectors). The final number of principal components participating in the construction of the individual observer variation feature matrix is ​​3M. ​​In addition to the bias principal components, the individual observer variation feature matrix also contains the average color matching function for the three channels. Each channel in the average color matching function also needs to be decomposed into a vector form to participate in the construction of the individual observer variation feature matrix. Therefore, the final individual observer variation feature matrix actually contains 3M+3 row vectors or column vectors. Assuming the color matching function is an XYZ color matching function, the first three principal components are retained for each channel, denoted as _____. , Representing the first 3 principal components of the x channel The other two are similar. The individual observer change feature matrix constructed at this point... It can be regarded as The matrix The dimension is The average color matching function for each channel and biased principal components They are all column vectors.

[0051] However, it should be noted that although the average color matching function and the principal component of the deviation for each channel in the above example are used to construct the individual observer variation feature matrix in column vector form, they can also be used in row vector form. This needs to be determined based on the specific projection calculation method of the projection matrix, prioritizing matching matrix dimensions. If they do not match, dimensional transformation can be performed through transposition.

[0052] It should be noted that the observer group data used in this invention to construct the individual observer variation feature matrix can be existing color matching function datasets or self-built color matching function datasets. These color matching function datasets can be actually measured color matching function data or computer-simulated color matching function data. A typical example of actually measured color matching function data is the 49 color matching function datasets from Stiles & Burch. A typical example of computer-simulated color matching function data is data simulated by Asano et al. based on the distribution of physiological visual parameters in a population. Such datasets randomly generated by the model are usually given in the form of physiological visual parameters, which typically include lens pigment optical density, macular pigment optical density, L, M, and S cone cell photosensitive pigment optical density, L and M cone cell spectral shifts, and the ability to generate LMS and XYZ color matching functions.

[0053] In this invention, the step of obtaining the individual observer change feature matrix in step S1 can be either by directly reading a pre-constructed individual observer change feature matrix or by performing the construction process of the individual observer change feature matrix online; there is no limitation on this. However, in practical applications, an individual observer change feature matrix can be pre-constructed and stored, and then directly read when optimizing for any color sample.

[0054] S2. For a color sample that needs to be reproduced on a multi-color display, the optimization objective is to minimize the projection error between the output spectrum of the color sample generated by the display driving vector on the multi-color display and the target spectrum of the color sample in the subspace defined by the individual observer variation feature matrix. Under the constraints of accurate color matching and display driving physical constraints, the display driving vector is optimized to obtain the optimal display driving vector that reduces the observer metamerism problem of the color sample.

[0055] It should be noted that the color sample to be reproduced on a multi-color display is a pure color sample, whose original reflectance spectrum is... At the target light source (spectral density denoted as ) The target spectrum under illumination is In practical applications, the color sample can be a standard color chart or an actual object, or it can be a specified three-channel value (e.g., a specified set of RGB values). However, the original reflectance spectrum of a standard color chart or an actual object can be measured, while the three-channel value needs to be converted using the spectral energy distribution of a standard light source combined with existing spectral reconstruction algorithms, lookup tables, or matrix transformations.

[0056] In an embodiment of the present invention, during the optimization process of the display driving vector, for any feasible solution of the display driving vector, the method for calculating the projection error in the objective function is as follows: The primary color spectrum P of the multi-primary color display is compared with the display driving vector corresponding to the current feasible solution. Multiplication yields the output spectrum of a multi-color display. Note the driving vector This refers to the driving value of a real display after linearizing and calibrating its gamma and other electro-optical transfer functions; that is, the intensity of the primary color as a function of the driving vector. Calculate the output spectrum based on linear variation. The target spectrum of the color sample under the target light source Deviation spectrum Then, the individual observer variation feature matrix W is used as the projection matrix to analyze the bias spectrum. The projection process can be expressed by the formula: The projection result is the projection error, which can then be used as the objective function value corresponding to the current feasible solution. This objective function aims to minimize the difference between the display's reproduced spectrum and the target spectrum in the average perception of the general population.

[0057] Furthermore, in other embodiments of the present invention, if it is necessary to reflect the differences in importance of different feature vectors (e.g., the average CMF is more important than higher-order principal components, and lower-order principal components are more important than higher-order principal components), then the objective function here can be weighted by the individual observer variation feature matrix W. This weighting can be achieved by introducing a diagonal matrix, and the objective function then takes the form: ,in It is a diagonal matrix, and the diagonal elements are the weights. If the dimension of the individual observer change feature matrix W is... Then the dimension of the diagonal matrix is .

[0058] In addition, the objective function solution in this invention also needs to follow certain constraints, which include two types of constraints: color accuracy matching constraints and display driving physical constraints.

[0059] The first type of chromaticity exact matching constraint uses an equality constraint, which takes the form: the tristimulus values ​​obtained by integrating the output spectrum and the target spectrum of the color sample using the CIE standard color matching function are equal. The formula for the chromaticity exact matching constraint can be expressed as follows: This ensures that the display output spectrum is within the CIE standard color matching function C. std The color must be precisely matched to the target spectrum, thus ensuring compatibility with existing color industry standards.

[0060] The second type of display driving physical constraint uses an inequality constraint, which takes the form: each element in the display driving vector of a multi-primary-color display is between 0 and the maximum driving value. The formula for the display driving physical constraint can be expressed as follows: , Let be the i-th element in the display driving vector, representing the driving value of the i-th primary color among the m primary colors of a multi-primary-color display. This represents the maximum driving value supported by a multi-color display. Generally, the driving values ​​for different primary colors of a display must be within the physically valid range, and are usually normalized to [value missing]. Therefore, after normalization, the physical constraint formula for display driving can be expressed as: .

[0061] In the above solution process, the sample data can be data obtained from actual measurements on color samples, including the primary color spectrum P of the multi-primary color display and the target spectrum of the color samples under the target light source. The optimization algorithm can be selected based on actual needs. In the embodiments of this invention, the above optimization problem can be formalized as a standard constrained quadratic programming (QP) problem, so the QP solver can be directly called to solve the display driving vector. optimal solution Perform the solution. After the solution is complete, the driving vector can be... It is used in multi-color displays to reproduce the target color of the corresponding color sample.

[0062] Additionally, in other embodiments of the invention, if the QP solver reports that the problem is infeasible, it indicates the target color. The tristimulus values ​​are located outside the display's color gamut. In this case, a gamut mapping step can be added before solving QP, selecting the closest point on the display's color gamut boundary.

[0063] This invention will further demonstrate the detailed implementation process and technical effects of the multi-primary color display optimization method based on individual observer characteristics shown in steps S1 to S2 above on specific data through several specific embodiments, so as to facilitate understanding of the essence of this invention.

[0064] Example 1

[0065] This embodiment provides a multi-primary-color display optimization method based on individual observer characteristics. This embodiment demonstrates color reproduction on a 5-primary-color display (the peak wavelengths of the 5 elements, from longest to shortest, are 702nm, 582nm, 521nm, 462nm, and 412nm) to showcase the specific technical effects of the invention. Figure 2 This demonstrates the normalized output spectrum of the display at maximum output power. Furthermore, this embodiment will be compared with a conventional 3-color display (RGB peak wavelengths of 636nm, 521nm, and 462nm, respectively) to illustrate the advantages of the present invention in reducing the observer metamerism index.

[0066] In this embodiment, the specific steps of the multi-primary-color display optimization method based on individual observer characteristics are as follows:

[0067] Step 1: Model and Data Preparation

[0068] Obtaining the color matching function dataset of the observer group: In this embodiment, the LMS color matching function data of 100 individual observers measured by Luo Ming's team at Zhejiang University are used. Each individual includes two classes: 2 degrees and 10 degrees. Figure 3 The LMS color matching function for 100 observers is shown.

[0069] Step 2: Construction of a feature model of color matching function changes for individual observers

[0070] A feature model describing the variation of the color matching function (CMF) among individual observers is obtained and constructed. This model is obtained by principal component analysis (PCA) on the CMF of a series of real individual observers. For the color matching functions at 2 degrees and 10 degrees for 100 individual observers in the dataset, the average CMF of all individual observers is first calculated: Since data from 2 degrees and 10 degrees are included, N=200 here. Additionally, the individual observer CMF here uses the LMS color matching function, therefore... The mean CMF for the L, M, and S channels is also included. Next, the CMF bias for each individual observer is calculated: Principal component analysis (PCA) was performed on the CMF bias of each color channel (L, M, S) for all observers, yielding the first M=3 principal components (eigenvectors) for each. When only the first 3 principal components are retained, the result is... . Figure 4 The average components of the L, M, and S channels were plotted. And the first three principal components. The individual observer change feature matrix constructed at this point. The dimension is K represents the number of discrete sampling wavelength points in the spectrum, and its matrix form is as follows: .

[0071] Step 3: Construct the constrained optimization problem

[0072] For each color sample to be reproduced, its original reflectance spectrum is: In the standard light source D65 (spectral density is...), The target spectrum under illumination is In this embodiment, a total of 1000 color samples were collected, with sampling intervals of 1 nm in the wavelength range of 390 nm to 780 nm. Figure 5 This is the spectrum of a color sample under a D65 light source, i.e., the target spectrum. Examples.

[0073] The goal of this embodiment is to find an optimal display driving vector. The vector needs to satisfy the following optimization objective and constraints:

[0074] 3.1: Optimization Objective (Objective Function)

[0075] In this embodiment, it is necessary to minimize the display output spectrum. With the target spectrum In the "Individual Observer Change Feature Matrix" The projection error in the defined subspace. Therefore, mathematically, we need to minimize the following objective function that aims to minimize the difference between the displayed spectrum and the target spectrum for the average perception of the general population:

[0076]

[0077] 3.2: Equality Constraints

[0078] In this embodiment, two types of constraints are set: chroma precision matching constraint and display driver physical constraint.

[0079] 1) Colorimetric precision matching constraint: The display output spectrum must be within the CIE standard color matching function C... std The color must precisely match the target spectrum, and mathematically, the following equation must be satisfied:

[0080]

[0081] This constraint ensures compatibility with existing color industry standards. The CIE standard color matching function C is used. std The CIE1931 standard color matching function can be used.

[0082] 2) Physical constraints of display driving

[0083] This constraint is an inequality constraint, requiring that the driving values ​​of different primary colors in a multi-primary-color display must be within the physically valid range. Since the driving values ​​are typically normalized... Therefore, mathematically, the following inequality must be satisfied:

[0084]

[0085] In the formula: is the i-th element in the display driving vector, representing the driving value of the i-th primary color in the multi-primary-color display; m is the total number of primary colors in the multi-primary-color display.

[0086] Step 4: Solving and Application

[0087] In this step, the constrained optimization problem is formalized into a standard constrained quadratic programming (QP) problem. For each color sample to be reproduced, based on the sample data of the color sample, the QP problem is solved using the numerical optimization solver (QP solver) in MATLAB to obtain the display driving vector. The optimal solution is denoted as the optimal driving vector. . This driving vector This is applied to multi-color displays to reproduce the target color of each color sample. Note that if the driving values ​​for different primary colors were normalized during optimization, then corresponding inverse normalization is required when applying this to a multi-color display.

[0088] It is important to note that the optimization here needs to be performed individually for each color sample to obtain the optimal driving vector. This is limited to the optimized display of that color sample. Therefore, in this embodiment, the same solution operation is performed on the aforementioned 1000 color samples in this step. Figure 6 Applying the optimal driving vector to a 5-primary-color display An example output spectrum is shown below.

[0089] Furthermore, in this embodiment, in order to evaluate the multi-primary color display optimization method based on individual observer characteristics proposed in this invention, according to... Figure 7 The process involves calculating the observer metamerism index to evaluate its performance. Using LMS color matching function data from 100 individual observers collected above, the observer metamerism was calculated. For a 3-color display, the average metamerism difference was calculated for each of the 100 observers across 1000 color samples. For these 100 individuals, the average CIEDE2000 color difference was 2.92, while for a corresponding 5-color display, the average CIEDE2000 color difference was 1.14, only about 39% of that of a traditional 3-color display. This demonstrates the advantage of the proposed method in improving the consistency of observer color perception.

[0090] Example 2

[0091] This embodiment demonstrates how the multi-color display optimization method based on individual observer characteristics proposed in this invention can be applied to a three-dimensional lookup table (3DLUT) in a practical industrial application of multi-color displays.

[0092] Although Example 1 demonstrates the advantage of this invention in reducing the observer metamerism index when processing the spectrum of a single color sample, in practical industrial applications of multi-primary-color displays (such as real-time video streaming), solving the high-precision spectral matching and quadratic programming (QP) problem for each pixel consumes a large amount of computational resources, making real-time computation difficult. Therefore, this embodiment further provides a specific industrial application method with low latency and low computational overhead, which combines offline computation with real-time interpolation to efficiently implement the optimization model of this invention. Specifically, it includes the following steps:

[0093] Step A: Establish an interpolation node network for the standard color space.

[0094] In a standard RGB color space (such as sRGB, where each channel value ranges from 0 to 255), a series of uniformly or non-uniformly distributed RGB interpolation nodes are created to construct a three-dimensional color space grid. For example, a 17×17×17 three-dimensional grid node array (with a total of 4913 nodes) can be created to cover the entire conventional RGB color space.

[0095] Step B: Node RGB conversion to target spectrum (preprocessing stage)

[0096] The 17×17×17 RGB interpolation nodes constructed above are converted into their respective target spectra (target reflection / emission spectra). The conversion from RGB nodes to target spectra here is a prior art in this field, and can be accomplished by combining the spectral energy distribution of a certain standard light source with existing spectral reconstruction algorithms, lookup tables, or matrix transformations.

[0097] Step C: Generate a multi-primary-color driven lookup table using the method of this invention (preprocessing and solving stage)

[0098] The spectrum corresponding to each node generated in step B is used as the target spectrum. The input is then fed into the core optimization method described in this invention (i.e., executing the aforementioned S1 and S2 steps, wherein when constructing the individual observer variation feature matrix W, the color matching function for all individual observers in the observer group should adopt the RGB color matching function). For each of these 17×17×17 nodes, a constrained optimization problem is established and solved using the individual observer variation feature matrix W, yielding the optimal display driving vector that minimizes observer metamerism and satisfies CIE standard chromaticity matching. Therefore, the original RGB input nodes are converted into the corresponding driving values ​​of n primary colors (n is the number of primary colors in a multi-primary-color display; for example, in a 5-primary-color display, n=5). Finally, a multi-primary-color display driver conversion data array of size (17×17×17)×n is generated and stored (i.e., each node stores an optimal display driving vector). ).

[0099] Step D: Real-time interpolation rendering based on RGB input source (real-time running stage)

[0100] During the actual operation of the display, when a new standard RGB video stream, image, or real-time rendering source is connected, the processing unit reads the real-time RGB value of each pixel, converts this RGB value into the corresponding target spectrum, and uses a pre-stored (17×17×17)×n three-dimensional lookup table to directly interpolate and calculate the final n primary color output driving value corresponding to the pixel through a fast spatial interpolation algorithm, and sends it to the display.

[0101] In this embodiment, the calculation of individual observer feature models is transferred to the offline pre-calculation stage before the display leaves the factory or during color calibration. In the real-time usage scenario of end consumers, only table lookup and interpolation operations with extremely low computational load are retained, which can significantly reduce the multi-primary color display optimization of observer metamerism with the advantages of low computing power requirements and high real-time performance.

[0102] Therefore, this embodiment provides a multi-primary-color display system based on individual observer characteristics, which includes:

[0103] The optimal driving vector storage module is used to store the target spectrum and the optimal display driving vector corresponding to the discrete color node in the three-dimensional color space, as a multi-primary color driving lookup table; the optimal display driving vector is pre-optimized by taking the discrete color node as the color sample to be reproduced on the multi-primary color display, according to the multi-primary color display optimization method based on individual observer characteristics as shown in S1 and S2 above.

[0104] The real-time rendering module is used to read the three-channel values ​​of each pixel that needs to be displayed in real time in the three-dimensional color space during the operation of the multi-primary color display, and convert them to obtain the target spectrum of the three-channel values ​​under the target light source. Then, based on the multi-primary color drive lookup table, the optimal display drive vector corresponding to this target spectrum is calculated by interpolation algorithm and sent to the display.

[0105] It should be noted that the method steps shown in S1~S2 above can essentially be implemented in the form of computer programs or software functional modules.

[0106] Therefore, based on the same inventive concept, such as Figure 8 As shown, the present invention also provides a computer electronic device corresponding to the multi-primary color display optimization method based on individual observer characteristics provided in the above embodiments, which includes a memory and a processor;

[0107] The memory is used to store computer programs;

[0108] The processor is configured to implement the multi-primary-color display optimization method based on individual observer characteristics as described above when executing the computer program.

[0109] Furthermore, the logical instructions in the aforementioned memory can be implemented as software functional units and, when sold or used as independent products, can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or a portion of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present invention.

[0110] Therefore, based on the same inventive concept, the present invention provides a computer-readable storage medium corresponding to the multi-primary color display optimization method based on individual observer characteristics. The storage medium stores a computer program, which, when executed by a processor, can realize the multi-primary color display optimization method based on individual observer characteristics as described above.

[0111] Therefore, based on the same inventive concept, the present invention provides a computer program product, including a computer program / instruction, which, when executed by a processor, can realize the multi-primary color display optimization method based on individual observer characteristics as described above.

[0112] Specifically, in the computer-readable storage medium of the above three embodiments, the stored computer program is executed by a processor, which can perform the aforementioned steps S1 to S2.

[0113] It is understood that the aforementioned storage media may include random access memory (RAM) or non-volatile memory (NVM), such as at least one disk storage device. Furthermore, the storage media may also be various media capable of storing program code, such as USB flash drives, external hard drives, magnetic disks, or optical discs.

[0114] It is understood that the processors mentioned above can be general-purpose processors, including central processing units (CPUs), network processors (NPs), etc.; they can also be digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, or discrete hardware components.

[0115] It should also be noted that those skilled in the art will understand that, for the sake of convenience and brevity, the specific working process of the system described above can be referred to the corresponding process in the foregoing method embodiments, and will not be repeated here. In the embodiments provided in this application, the division of steps or modules in the system and method is merely a logical functional division, and there may be other division methods in actual implementation. For example, multiple modules or steps may be combined or integrated together, and a module or step may also be split.

[0116] The embodiments described above are merely some preferred implementations of the present invention and are not intended to limit the invention. Those skilled in the art can make various changes and modifications without departing from the spirit and scope of the invention. Therefore, all technical solutions obtained through equivalent substitution or transformation fall within the protection scope of the present invention.

Claims

1. A multi-primary-color display optimization method based on individual observer characteristics, characterized in that, include: S1. Obtain the individual observer variation feature matrix composed of the average color matching function and the bias principal components; the average color matching function is obtained by averaging the color matching functions of the observer group; the bias principal components are obtained by principal component analysis of the bias functions of the color matching functions of all individual observers in the observer group relative to the average color matching function; S2. For a color sample that needs to be reproduced on a multi-color display, the optimization objective is to minimize the projection error between the output spectrum of the color sample generated by the display driving vector on the multi-color display and the target spectrum of the color sample in the subspace defined by the individual observer variation feature matrix. Under the constraints of accurate color matching and display driving physical constraints, the display driving vector is optimized to obtain the optimal display driving vector.

2. The multi-primary color display optimization method based on individual observer characteristics as described in claim 1, characterized in that, The color matching function of each individual observer in the observer group is of the same type, selected from one of the LMS color matching function, XYZ color matching function, or RGB color matching function.

3. The multi-primary-color display optimization method based on individual observer characteristics as described in claim 1, characterized in that, The average color matching function and the bias principal component each contain three color channels. The average color matching function of each color channel and the bias principal component of each color channel are combined in the form of row vectors or column vectors to form the individual observer change feature matrix.

4. The multi-primary color display optimization method based on individual observer characteristics as described in claim 1, characterized in that, The method for obtaining the bias principal components is as follows: For the observer group, the color matching function of each individual observer is subtracted from the average color matching function to obtain the three-channel bias function of each individual observer; principal component analysis is performed on the three-channel bias function of all individual observers in the observer group channel by channel, and a preset number of principal components are extracted for each color channel to be used as bias principal components to construct the individual observer change feature matrix.

5. The multi-primary-color display optimization method based on individual observer characteristics as described in claim 1, characterized in that, In the process of optimizing the display driving vector, for any feasible solution of the display driving vector, the method for calculating the projection error in the objective function is as follows: multiply the primary color spectrum of the multi-primary color display with the display driving vector corresponding to the current feasible solution to obtain the output spectrum of the multi-primary color display, calculate the deviation spectrum between the output spectrum and the target spectrum of the color sample under the target light source, and then use the individual observer change feature matrix as the projection matrix to project the deviation spectrum. The resulting projection error is used as the objective function value corresponding to the current feasible solution.

6. The multi-primary color display optimization method based on individual observer characteristics as described in claim 1, characterized in that, The colorimetric precision matching constraint adopts an equality constraint, which is in the form that the tristimulus values ​​obtained by integrating the output spectrum and the target spectrum of the color sample through the CIE standard color matching function are equal.

7. The multi-primary color display optimization method based on individual observer characteristics as described in claim 1, characterized in that, The physical constraints for driving the display adopt inequality constraints, which are in the form that each element in the display driving vector of the multi-primary-color display is between 0 and the maximum driving value.

8. A multi-primary-color display system based on individual observer characteristics, characterized in that, include: The optimal driving vector storage module is used to store the target spectrum and the optimal display driving vector corresponding to the discrete color nodes in the three-dimensional color space, as a multi-primary color driving lookup table. The optimal display driving vector is obtained by pre-optimizing the discrete color node as the color sample to be reproduced on the multi-primary color display according to any one of claims 1 to 7, using the multi-primary color display optimization method based on individual observer features as the discrete color node in advance. The real-time rendering module is used to read the three-channel values ​​of each pixel that needs to be displayed in real time in the three-dimensional color space during the operation of the multi-primary color display, and convert them to obtain the target spectrum of the three-channel values ​​under the target light source. Then, based on the multi-primary color drive lookup table, the optimal display drive vector corresponding to this target spectrum is calculated by interpolation algorithm and sent to the display.

9. A computer program product comprising a computer program / instructions, characterized in that, When the computer program / instruction is executed by the processor, it can implement the multi-primary color display optimization method based on individual observer characteristics as described in any one of claims 1 to 7.

10. A computer electronic device, characterized in that, Including memory and processor; The memory is used to store computer programs; The processor is configured to, when executing the computer program, implement the multi-primary color display optimization method based on individual observer characteristics as described in any one of claims 1 to 7.

Images