CN117690388B - Picture optimization method and system based on display module backlight - Google Patents
Picture optimization method and system based on display module backlight Download PDFInfo
- Publication number
- CN117690388B CN117690388B CN202410155676.1A CN202410155676A CN117690388B CN 117690388 B CN117690388 B CN 117690388B CN 202410155676 A CN202410155676 A CN 202410155676A CN 117690388 B CN117690388 B CN 117690388B
- Authority
- CN
- China
- Prior art keywords
- optimization
- backlight
- target
- backlight module
- value
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
- 238000005457 optimization Methods 0.000 title claims abstract description 337
- 238000000034 method Methods 0.000 title claims abstract description 59
- 238000001228 spectrum Methods 0.000 claims abstract description 129
- 238000004458 analytical method Methods 0.000 claims abstract description 32
- 230000010354 integration Effects 0.000 claims abstract description 26
- 238000012360 testing method Methods 0.000 claims abstract description 24
- 238000010183 spectrum analysis Methods 0.000 claims abstract description 12
- 238000009826 distribution Methods 0.000 claims description 93
- 238000004364 calculation method Methods 0.000 claims description 59
- 230000006870 function Effects 0.000 claims description 56
- 230000009471 action Effects 0.000 claims description 49
- 238000004422 calculation algorithm Methods 0.000 claims description 43
- 230000003595 spectral effect Effects 0.000 claims description 41
- 230000000694 effects Effects 0.000 claims description 31
- 230000008569 process Effects 0.000 claims description 23
- 238000013528 artificial neural network Methods 0.000 claims description 19
- 238000013041 optical simulation Methods 0.000 claims description 19
- 230000003287 optical effect Effects 0.000 claims description 14
- 238000011156 evaluation Methods 0.000 claims description 10
- 230000000007 visual effect Effects 0.000 claims description 9
- 230000008901 benefit Effects 0.000 claims description 8
- 238000013433 optimization analysis Methods 0.000 claims description 6
- 238000005315 distribution function Methods 0.000 claims description 5
- 238000012545 processing Methods 0.000 claims description 4
- 230000004044 response Effects 0.000 claims description 4
- 230000032683 aging Effects 0.000 claims description 3
- 230000001902 propagating effect Effects 0.000 claims description 3
- 238000005316 response function Methods 0.000 claims description 3
- 230000035945 sensitivity Effects 0.000 claims description 3
- 230000007704 transition Effects 0.000 claims description 3
- 230000001186 cumulative effect Effects 0.000 claims description 2
- 230000001131 transforming effect Effects 0.000 claims description 2
- 238000013473 artificial intelligence Methods 0.000 abstract description 2
- 238000005265 energy consumption Methods 0.000 description 11
- 239000003086 colorant Substances 0.000 description 8
- 238000013461 design Methods 0.000 description 8
- 230000035772 mutation Effects 0.000 description 6
- 230000007774 longterm Effects 0.000 description 5
- 230000006399 behavior Effects 0.000 description 4
- 230000007246 mechanism Effects 0.000 description 3
- 238000004088 simulation Methods 0.000 description 3
- 238000003860 storage Methods 0.000 description 3
- 238000010586 diagram Methods 0.000 description 2
- 230000007614 genetic variation Effects 0.000 description 2
- 230000017525 heat dissipation Effects 0.000 description 2
- 238000003491 array Methods 0.000 description 1
- 238000009395 breeding Methods 0.000 description 1
- 230000001488 breeding effect Effects 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 230000002301 combined effect Effects 0.000 description 1
- 230000001419 dependent effect Effects 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000018109 developmental process Effects 0.000 description 1
- 230000009977 dual effect Effects 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 238000005286 illumination Methods 0.000 description 1
- 239000004973 liquid crystal related substance Substances 0.000 description 1
- 238000007726 management method Methods 0.000 description 1
- 239000000463 material Substances 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000009877 rendering Methods 0.000 description 1
- 238000012216 screening Methods 0.000 description 1
- 238000000638 solvent extraction Methods 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
- 238000001429 visible spectrum Methods 0.000 description 1
Abstract
The application relates to the technical field of artificial intelligence and discloses a picture optimization method and a system based on display module backlight. The method comprises the following steps: creating a backlight module model and collecting a plurality of target backlight module parameters; performing picture optimization execution strategy analysis to obtain an initial picture optimization execution strategy; carrying out backlight spectrum analysis and linear programming solution to obtain initial backlight spectrum optimization parameters; performing local dimming optimization to obtain target backlight spectrum optimization parameters; performing LED array analysis and Fourier series optimization on the backlight module model based on an initial picture optimization execution strategy to obtain target LED array optimization parameters; and carrying out strategy optimization and module integration test on the initial picture optimization execution strategy according to the target backlight spectrum optimization parameters and the target LED array optimization parameters to obtain a target picture optimization execution strategy, thereby improving the picture optimization accuracy of the backlight of the display module.
Description
Technical Field
The application relates to the technical field of artificial intelligence, in particular to a picture optimization method and system based on display module backlight.
Background
Traditional backlight module designs often employ a fixed illumination scheme, which makes it difficult to achieve an optimal display effect in a specific use scenario, for example, uneven brightness or color distortion easily occurs in a high-contrast scenario, and affects the visual experience of a user and the energy efficiency of equipment.
In addition, with the rapid development of display technology, the design and optimization of the backlight module become increasingly complex. From the layout of the LED array, the material selection and structural design of the optical elements, to the control of power management and heat distribution, each link has a direct influence on the final display effect of the picture. These factors are often dependent and influence each other, and it is difficult for a single optimization strategy to meet the design requirements of multiple objectives and high efficiency. Therefore, how to achieve cost-effective and energy-consuming optimization while guaranteeing picture quality has become a significant challenge in this field. Conventional optimization methods often rely on empirical design and item-by-item adjustments, which are not only inefficient, but also difficult to accommodate with increasingly diverse application requirements.
Disclosure of Invention
The application provides a picture optimization method and a system based on a display module backlight, which further improve the picture optimization accuracy of the display module backlight.
The first aspect of the application provides a picture optimization method based on a display module backlight, which comprises the following steps:
Creating a backlight module model through LightTools software and module performance constraint conditions, and performing optical simulation and backlight module parameter acquisition on the backlight module model through MATLAB software to obtain a plurality of target backlight module parameters;
Defining a state space and an action space of the backlight module model according to the target backlight module parameters, and performing picture optimization execution strategy analysis on the backlight module model through a double-depth Q network algorithm to obtain an initial picture optimization execution strategy;
performing backlight spectrum analysis on the backlight module model based on the initial picture optimization execution strategy to obtain spectrum distribution data, and performing linear programming solution on the spectrum distribution data to obtain initial backlight spectrum optimization parameters;
carrying out local dimming optimization on the initial backlight spectrum optimization parameters through a preset MOEA/D-SFLA algorithm to obtain target backlight spectrum optimization parameters;
Performing LED array analysis and Fourier series optimization on the backlight module model based on the initial picture optimization execution strategy to obtain target LED array optimization parameters;
And carrying out strategy optimization and module integration test on the initial picture optimization execution strategy according to the target backlight spectrum optimization parameters and the target LED array optimization parameters to obtain a target picture optimization execution strategy.
The second aspect of the present application provides a screen optimization system based on a display module backlight, the screen optimization system based on the display module backlight comprising:
The system comprises a creation module, a module selection module and a module performance constraint condition module, wherein the creation module is used for creating a backlight module model through the LightTools software and the module performance constraint condition, and carrying out optical simulation and backlight module parameter acquisition on the backlight module model through MATLAB software to obtain a plurality of target backlight module parameters;
The definition module is used for defining a state space and an action space of the backlight module model according to the target backlight module parameters, and carrying out picture optimization execution strategy analysis on the backlight module model through a double-depth Q network algorithm to obtain an initial picture optimization execution strategy;
The solving module is used for carrying out backlight spectrum analysis on the backlight module model based on the initial picture optimization execution strategy to obtain spectrum distribution data, and carrying out linear programming solving on the spectrum distribution data to obtain initial backlight spectrum optimization parameters;
The processing module is used for carrying out local dimming optimization on the initial backlight spectrum optimization parameters through a preset MOEA/D-SFLA algorithm to obtain target backlight spectrum optimization parameters;
the analysis module is used for carrying out LED array analysis and Fourier series optimization on the backlight module model based on the initial picture optimization execution strategy to obtain target LED array optimization parameters;
And the optimization module is used for carrying out strategy optimization and module integration test on the initial picture optimization execution strategy according to the target backlight spectrum optimization parameter and the target LED array optimization parameter to obtain a target picture optimization execution strategy.
In the technical scheme provided by the application, by combining the advanced optical simulation of the LightTools and MATLAB software with the double-depth Q network (DDQN) algorithm and the MOEA/D-SFLA algorithm, the method not only focuses on single picture quality or energy efficiency, but also realizes the comprehensive optimization of light intensity distribution, spectrum distribution, energy consumption, cost and other multidimensional degrees. This allows the final display effect to reach an optimal balance in terms of brightness uniformity, color realism, energy efficiency, etc. By using the LightTools software to accurately model the backlight module and combining MATLAB to perform optical simulation and parameter acquisition, the propagation, scattering and reflection behaviors of light rays in the backlight module can be analyzed in detail, and accurate input data can be provided for subsequent optimization. This accurate modeling and analysis is the key basis for optimizing the effect. By applying an advanced double-depth Q network algorithm, an optimal action strategy is automatically learned and predicted, the brightness of the LED and the structure of an optical element can be intelligently adjusted, and different display requirements and use scenes can be dynamically adapted. This intelligent strategy greatly improves the efficiency and effectiveness of the optimization. By combining with the MOEA/D-SFLA algorithm, the problems of global optimization and local dimming optimization can be effectively solved, the local area can be finely adjusted while the optimal solution is ensured to be found in the global range, and the picture quality is further improved. Not only is the optimization performed at the component level, but also the integration and coordination of the entire backlight module system is concerned. Through simulation and integration test of a system level, the cooperative work of each component and strategy is ensured, and further the picture optimization accuracy of the backlight of the display module is improved.
Drawings
FIG. 1 is a diagram illustrating an embodiment of a method for optimizing a frame based on a backlight of a display module according to an embodiment of the present application;
Fig. 2 is a schematic diagram of an embodiment of a screen optimization system based on a backlight of a display module according to an embodiment of the application.
Detailed Description
The embodiment of the application provides a picture optimization method and a system based on a display module backlight, which further improve the picture optimization accuracy of the display module backlight.
The terms "first," "second," "third," "fourth" and the like in the description and in the claims and in the above drawings, if any, are used for distinguishing between similar objects and not necessarily for describing a particular sequential or chronological order. It is to be understood that the data so used may be interchanged where appropriate such that the embodiments described herein may be implemented in other sequences than those illustrated or otherwise described herein. Furthermore, the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, system, article, or apparatus that comprises a list of steps or elements is not necessarily limited to those steps or elements expressly listed or inherent to such process, method, article, or apparatus.
For easy understanding, the following describes a specific flow of an embodiment of the present application, referring to fig. 1, and an embodiment of a method for optimizing a screen based on a backlight of a display module according to the embodiment of the present application includes:
Step 101, creating a backlight module model through LightTools software and module performance constraint conditions, and performing optical simulation and backlight module parameter acquisition on the backlight module model through MATLAB software to obtain a plurality of target backlight module parameters;
It can be understood that the execution subject of the present application may be a screen optimization system based on a backlight of a display module, and may also be a terminal or a server, which is not limited herein. The embodiment of the application is described by taking a server as an execution main body as an example.
Specifically, a backlight module model is created by the LightTools software according to module performance constraint conditions, wherein the constraint conditions comprise power consumption limitation, cost budget and size limitation, so that the model is ensured to be in line with an actual application scene and feasible in cost and size. The created model comprises N backlight modules, and each module is a basic unit of the optimization process. And (3) performing optical simulation on the backlight module model by using MATLAB software, and performing accurate optical characteristic analysis on the model. In the optical simulation process, light intensity calculation is performed, and the initial light intensity, the light attenuation coefficient, the distance between the light and the observation point and the included angle between the light and the normal line of each LED unit are considered, so that the light intensity of any point in the three-dimensional space is accurately estimated. The total power consumption calculation is performed on the backlight module model, which involves the working voltage, current and aging coefficient of each LED, and these parameters together determine the total power consumption of the backlight module. The power consumption calculation is helpful for evaluating the energy efficiency performance of the module and provides a reference for the subsequent energy consumption optimization. And (3) calculating the spectral efficiency, and evaluating the light output effect under unit power consumption by analyzing the spectral distribution of the backlight module under different wavelengths and the light efficiency of the wavelength corresponding to human eyes. The spectrum efficiency calculation is important to improving the energy efficiency and visual experience of the backlight module. And (3) performing color gamut coverage calculation, analyzing the spectrum distribution of different colors in a wavelength range, the weight of each color and the target color gamut coverage, ensuring that the backlight module can cover a wide color gamut, and providing richer and accurate color expression for display equipment. Finally, generating a plurality of target backlight module parameters of the backlight module model.
Step 102, defining a state space and an action space of a backlight module model according to a plurality of target backlight module parameters, and performing picture optimization execution strategy analysis on the backlight module model through a double-depth Q network algorithm to obtain an initial picture optimization execution strategy;
Specifically, a state space of a backlight module model is defined according to a plurality of target backlight module parameters, wherein the state space comprises light intensity distribution factors and spectrum distribution factors, and the factors directly influence the quality and consistency of a display effect. And defining an action space of a backlight module model according to a plurality of target backlight module parameters, wherein the action space comprises brightness adjustment of LEDs and structural adjustment of optical elements, and the actions are key means for optimizing the display effect. And inputting a plurality of target backlight module parameters of the backlight module model into a dual-depth Q network algorithm, wherein the algorithm comprises two depth neural networks which are respectively used for estimating the Q value of the current state and updating the Q value. And estimating the Q value of the current state of the backlight module model through the first deep neural network to obtain the Q value of the current state. The potential value of taking different actions in the current state is analyzed. And the second deep neural network updates the Q value of the current state to obtain the Q value of the next state. The Q value updating formula is a behavior-based value iterative method, wherein the learning rate determines the degree to which new information covers old information; the instant rewards reflect the direct effect of the current action; the discount factor is used to measure the importance of future rewards. Through this process, the algorithm is able to learn and predict the best action selection based on the current state and the outcome of the action. And calculating a target Q value according to the Q value of the next state, and evaluating the expected long-term benefit after taking a specific action in a specific state. The target Q calculation function takes into account the present values of the instant and future rewards, thereby providing a quantified target for the algorithm to guide the learning process. And performing bonus function calculation according to the target Q value, evaluating different action effects, and further analyzing a picture optimization execution strategy of the backlight module model by using the bonus function calculation result to finally obtain an initial picture optimization execution strategy.
Step 103, carrying out backlight spectrum analysis on the backlight module model based on an initial picture optimization execution strategy to obtain spectrum distribution data, and carrying out linear programming solution on the spectrum distribution data to obtain initial backlight spectrum optimization parameters;
It should be noted that, based on the initial image optimization execution strategy, the backlight module model is subjected to spectrum analysis to obtain spectrum distribution data. The data provides detailed information about the spectral intensities at different wavelengths, which is the basis for optimizing the performance of the backlight module. And constructing a linear programming model of the backlight module model, maximizing the luminous efficiency of the backlight module by a mathematical optimization method, and simultaneously ensuring sufficient color gamut coverage. In this process, one core goal of the linear programming model is to optimize the luminous efficiency, i.e. to reduce the energy consumption as much as possible, while guaranteeing the display effect. To achieve this objective, an objective function of luminous efficiency optimization is used that calculates luminous efficiency by quantifying the spectral intensity at the wavelengths and the relative visual response of the human eye to these wavelengths of light, in comparison to the total power consumption. In this process, it is critical to balance the display quality and the energy consumption to achieve an efficient and economical display effect. Meanwhile, the linear programming model also comprises constraint conditions of color gamut coverage. These conditions ensure that the backlight module can cover the necessary color gamut range while optimizing the luminous efficiency, and ensure the authenticity and richness of the display effect. The calculation of the gamut coverage optimization parameters involves analyzing the spectral response functions of the different colors at a specific wavelength and the minimum coverage requirements of these colors. Therefore, the backlight module can still provide wide and accurate color expression while reducing energy consumption. And combining the luminous efficiency optimization parameter and the color gamut coverage optimization parameter to generate an initial backlight spectrum optimization parameter of the backlight module model. These parameters are key outputs of the backlight module optimization process, and they directly affect the overall performance of the display module, including brightness, color rendition, energy efficiency, etc.
104, Performing local dimming optimization on the initial backlight spectrum optimization parameters through a preset MOEA/D-SFLA algorithm to obtain target backlight spectrum optimization parameters;
Specifically, the initial backlight spectrum optimization parameters are subjected to population initialization through a MOEA/D-SFLA algorithm, and a plurality of first backlight spectrum optimization parameters are generated. And evaluating each first backlight spectrum optimization parameter through a preset fitness function, and calculating first fitness values of the first backlight spectrum optimization parameters. These fitness values are important indicators that measure how each optimization parameter behaves at the current objective, and they will be compared to the set first and second objective values. The first target value and the second target value set different performance criteria, the first target value being smaller than the second target value reflecting different desired levels of the optimization parameter. And dividing the first backlight spectrum optimization parameter into different groups according to the comparison result of the first fitness value and the two target values. If the first fitness value is smaller than the first target value, dividing the corresponding parameters into a first spectrum optimization parameter group; if the first fitness value is between the first target value and the second target value, partitioning into a second population of spectrally optimized parameters; if the second target value is greater than the second target value, dividing into a third spectrum optimization parameter group. The classification mechanism ensures that parameters of different performance levels are properly grouped, providing direction for subsequent optimization. The first spectrum optimization parameter group and the second spectrum optimization parameter group are subjected to propagation, crossover and mutation operations, and the third spectrum optimization parameter group is subjected to crossover and mutation only. By simulating the mechanisms of natural selection and genetic variation, the evolution of parameters to a better solution is promoted. The breeding, crossover and mutation operations can create diversity, exploring a wider solution space. And calculating a second fitness value for each second backlight spectrum optimization parameter, and performing optimization analysis based on the fitness values. The spectral optimization parameter that performs best, i.e., the target backlight spectral optimization parameter, is identified.
Step 105, performing LED array analysis and Fourier series optimization on the backlight module model based on an initial picture optimization execution strategy to obtain target LED array optimization parameters;
Specifically, the LED array analysis is performed on the backlight module model based on the initial picture optimization execution strategy, and the layout information of the current LED array is obtained. According to the LED array layout information, light intensity distribution calculation is carried out on the backlight module model, a Fourier series method is adopted, and the change of light intensity at different positions is accurately described in a mathematical mode through converting the space distribution of the LED array into a frequency domain. In this calculation, each fourier coefficient corresponds to a preset frequency component, which is responsible for describing the specific pattern of the LED array in the spatial light intensity distribution. By calculating these coefficients, the distribution of light intensity in the LED array is understood and optimized. The light intensity distribution calculation function takes into account the frequency index along the x and y directions, as well as the dimensions of the LED array in both directions. The complex exponential function serves to represent the spatially varying modes in the light intensity distribution. And optimizing the light intensity distribution of the current LED array by utilizing a Fourier series method, and determining optimization parameters of the target LED array, including the optimal LED layout and angular intensity distribution. The magnitude of the contribution of the different frequency components to the light intensity distribution is accurately calculated by converting the light intensity distribution from the spatial domain to the frequency domain using a fourier series function. In this process, the calculation of the fourier coefficients represents how the contributions of each fourier coefficient are integrated over the entire LED array area. By calculation and optimization, an optimized LED array layout is obtained, which not only improves the overall quality of the display effect, but also optimizes the energy efficiency and uniformity of the light source.
And 106, performing strategy optimization and module integration test on the initial picture optimization execution strategy according to the target backlight spectrum optimization parameters and the target LED array optimization parameters to obtain a target picture optimization execution strategy.
Specifically, the initial picture optimization execution strategy is subjected to strategy optimization according to the target backlight spectrum optimization parameters and the target LED array optimization parameters, and the initial strategy is adjusted and perfected so as to be more in line with actual display requirements and performance indexes. And generating a plurality of candidate picture optimization execution strategies through strategy optimization. And carrying out module integration test on the candidate picture optimization execution strategies, and verifying and evaluating the actual effect of each strategy. Each candidate strategy is tested through the backlight module model, and detailed data about the performance of the strategies in practical application is collected. The module integrated test data not only comprises the effect of the strategy on improving the picture quality, but also covers performance indexes such as energy consumption, stability and the like. And carrying out strategy evaluation on each candidate picture optimization execution strategy according to the module integration test data. The performance of each strategy was quantified and their combined effect was assessed accordingly. The policy evaluation indexes comprise the degree of improving the picture quality, the energy efficiency ratio, the long-term stability and the like, and the indexes comprehensively reflect the advantages and disadvantages and the applicability of each policy. And carrying out optimization solving on the multiple candidate picture optimization execution strategies according to the strategy evaluation indexes. This step is the key in the whole process to screen out the optimal picture optimization execution strategy from all candidate strategies. And comprehensively considering various indexes to find out strategies with optimal performance in aspects of picture quality, energy efficiency, stability and the like. By analysis and evaluation, the finally determined target picture optimization execution strategy can optimize the energy consumption and improve the performance of the whole display system while ensuring the picture quality. The display effect is improved, the sustainability and the practicability of the system are considered, and the high-efficiency and stable operation in practical application is ensured.
In the embodiment of the application, by combining the advanced optical simulation of the LightTools and MATLAB software with the double-depth Q network (DDQN) algorithm and the MOEA/D-SFLA algorithm, not only is single picture quality or energy efficiency focused, but also the comprehensive optimization of light intensity distribution, spectral distribution, energy consumption, cost and other multidimensional degrees is realized. This allows the final display effect to reach an optimal balance in terms of brightness uniformity, color realism, energy efficiency, etc. By using the LightTools software to accurately model the backlight module and combining MATLAB to perform optical simulation and parameter acquisition, the propagation, scattering and reflection behaviors of light rays in the backlight module can be analyzed in detail, and accurate input data can be provided for subsequent optimization. This accurate modeling and analysis is the key basis for optimizing the effect. By applying an advanced double-depth Q network algorithm, an optimal action strategy is automatically learned and predicted, the brightness of the LED and the structure of an optical element can be intelligently adjusted, and different display requirements and use scenes can be dynamically adapted. This intelligent strategy greatly improves the efficiency and effectiveness of the optimization. By combining with the MOEA/D-SFLA algorithm, the problems of global optimization and local dimming optimization can be effectively solved, the local area can be finely adjusted while the optimal solution is ensured to be found in the global range, and the picture quality is further improved. Not only is the optimization performed at the component level, but also the integration and coordination of the entire backlight module system is concerned. Through simulation and integration test of a system level, the cooperative work of each component and strategy is ensured, and further the picture optimization accuracy of the backlight of the display module is improved.
In a specific embodiment, the process of executing step 101 may specifically include the following steps:
(1) Obtaining module performance constraint conditions, wherein the module performance constraint conditions comprise: power consumption limitations, cost budgets, and size limitations;
(2) Creating a backlight module model according to module performance constraint conditions by using LightTools software, wherein the backlight module model comprises N backlight modules;
(3) Optical simulation is carried out on the backlight module model through MATLAB software, and light intensity calculation is carried out on the backlight module model to obtain light intensity, wherein the light intensity calculation function is as follows: ，/> coordinates in three-dimensional space Light intensity at,/>Represents the/>Initial light intensity of individual LEDs,/>Represents the/>The light attenuation coefficient of the individual LEDs,From the/>Individual LED to position/>Distance of/>From the/>Individual LEDs to positionThe included angle between the light rays and the normal line, N represents the total number of LEDs;
(4) And performing backlight module power consumption calculation on the backlight module model to obtain the total power consumption of the backlight module, wherein the backlight module power consumption calculation function is as follows: ，/> Indicating the total power consumption of the backlight module, First/>Operating voltage of individual LEDs,/>Represents the/>Operating current of individual LEDs,/>Represents the/>Aging coefficient of individual LEDs,/>Indicating the operating time of the LED;
(5) And performing spectral efficiency calculation on the backlight module model to obtain the spectral efficiency of the backlight module, wherein the spectral efficiency calculation function is as follows: ，/> the spectral efficiency of the backlight module is expressed as the light output effect per unit power consumption,/> Expressed in wavelength/>Spectral distribution function at,/>Representing the corresponding wavelength of the human eye/>Describing the sensitivity of the human eye to light of different wavelengths;
(6) Performing color gamut coverage calculation on the backlight module model to obtain color gamut coverage, wherein the color gamut coverage calculation function is as follows: ，/> For color gamut coverage, represent the ability of the backlight module to cover the target color gamut,/> Color/>At wavelength/>Spectral distribution function at,/>Representing color/>Representing the importance of color c in the overall gamut,/>Representing color/>Target coverage of/>Representing a gamut coverage function describing contributions of different wavelengths of light to the gamut coverage;
(7) And generating a plurality of target backlight module parameters of the backlight module model according to the light intensity, the total power consumption of the backlight module, the spectral efficiency of the backlight module and the color gamut coverage.
In particular, module performance constraints, including power consumption constraints, cost budget, and size constraints, are obtained that together define the framework for backlight module design. For example, a large outdoor display screen has higher power consumption limitations and size requirements, while a portable device has more stringent power consumption and size limitations. These constraints directly affect the design and performance of the backlight module. And creating a backlight module model according to the module performance constraint conditions through the LightTools software. Setting the arrangement mode, the quantity and the mutual position relation with other optical elements of the LEDs to form an integral model comprising N backlight modules. And carrying out optical simulation on the backlight module model by MATLAB software, and calculating the light intensity distribution. The light intensity calculation determines the brightness and uniformity of the display screen. In this process, the calculation function can take into account the initial light intensity, the light attenuation coefficient, the distance from the observation point, the included angle, and the like of each LED. For example, if one LED is farther from the center of the screen, its brightness contribution to the center area will decrease accordingly. And performing power consumption calculation on the backlight module model to determine the total power consumption of the backlight module. The power consumption calculation not only relates to the energy consumption of the module, but also influences the heat dissipation design and the long-term operation reliability. For example, an excessively high power consumption backlight module requires additional heat dissipation measures. Spectral efficiency calculations were performed. The spectral efficiency is a key index for measuring the light output effect of the backlight module, reflects the light output effect under unit power consumption, and needs to consider the spectral distribution of light rays with different wavelengths and the sensitivity of human eyes to the light rays with the wavelengths. For example, for green light that is more sensitive to the human eye, the backlight module needs to provide a sufficient green light output to ensure the nature and comfort of the overall display effect. And performing color gamut coverage calculation. The color gamut coverage is an indicator for evaluating the ability of the backlight module to cover the target color gamut. The color gamut coverage is directly related to how many colors the display can exhibit, as well as the saturation and accuracy of these colors. And determining the color gamut coverage of the backlight module by calculating the spectral distribution and the corresponding weight of different colors in the wavelength range.
In a specific embodiment, the process of executing step 102 may specifically include the following steps:
(1) Defining a state space of a backlight module model according to a plurality of target backlight module parameters, wherein the state space is as follows: a light intensity distribution factor and a spectral distribution factor;
(2) Defining an action space of a backlight module model according to a plurality of target backlight module parameters, wherein the action space comprises: brightness adjustment of the LEDs and structural adjustment of the optical elements;
(3) Inputting a plurality of target backlight module parameters of the backlight module model into a double-depth Q network algorithm, wherein the double-depth Q network algorithm comprises a first depth neural network and a second depth neural network;
(4) The method comprises the steps of estimating the Q value of the current state of a backlight module model through a first deep neural network to obtain the Q value of the current state, and updating the Q value of the current state of the backlight module model through a second deep neural network to obtain the Q value of the next state, wherein a Q value updating formula is as follows: ，/> Expressed in state/> Down execution action/>When according to policy/>Expected benefit obtained,/>For learning rate, determining the degree to which the newly received information covers the old information,/>Representing instant rewards, reflecting the direct effect of the current action,/>Representing discount factors, for measuring importance of future rewards,/>Representing the current state,/>Representing the next state,/>Representing the current action,/>Representing the next action,/>Representing network parameters,/>Is the current value,/>Is an old value;
(5) And calculating a target Q value of the backlight module model according to the Q value of the next state to obtain the target Q value, wherein the target Q value calculation function is as follows: ，/> represents the/> Target Q value of individual state-action pair,/>Represents the/>Instant rewards earned by individual state-action pairs,/>Representing a discount factor, representing the present value of a future reward,/>Representing the next state to transition to after an action is performed,/>Representing network parameters,/>Is the current value,/>Is an old value;
(6) And performing bonus function calculation on the target Q value to obtain a bonus function calculation result, and performing picture optimization execution strategy analysis on the backlight module model according to the bonus function calculation result to obtain an initial picture optimization execution strategy.
Specifically, a state space of a backlight module model is defined according to a plurality of target backlight module parameters, wherein the state space comprises light intensity distribution factors and spectrum distribution factors. For example, the state space includes brightness levels and color rendering capabilities of different LED areas. These factors together determine the visual effects of the display, such as brightness uniformity and color accuracy. Defining the action space of the backlight module model, including brightness adjustment of the LEDs, structure adjustment of the optical elements, and the like. For example, the action space may include adjusting the brightness of a particular LED area or changing the angle or position of certain optical elements to improve light distribution or reduce light loss. And inputting a plurality of target backlight module parameters of the backlight module model into a dual-depth Q network algorithm for processing. This algorithm includes two core parts: the first deep neural network is used for Q value estimation of the current state, and the second deep neural network is responsible for Q value update. With this structure, the algorithm can more accurately predict the long-term impact of each action. In the first deep neural network, the Q value of the current state of the backlight module model is estimated, and the Q value represents expected benefits brought by taking specific actions in the current state. For example, if the current state is that the brightness of a particular region is below an ideal level, increasing the brightness of the LEDs in that region increases the Q value of the state. And the second deep neural network updates the Q value of the current state to obtain the Q value of the next state. The Q value update formula combines the instant rewards and the estimation of future rewards, providing a way for the algorithm to balance the direct impact and long term effects of the current action. For example, increasing the LED brightness in a certain area immediately improves the picture quality (instant rewards), but also increases the power consumption (affects future rewards). Further, based on the Q value of the next state, target Q value calculation is performed on the backlight module model. The expected revenue for the next state that is transferred to after the specific action is performed is evaluated. The target Q calculation function combines the present values of the instant and future rewards to provide a quantized target for each state-action pair. And carrying out rewarding function calculation on the target Q value, and carrying out picture optimization execution strategy analysis on the backlight module model according to calculation results. The performance of different strategies in practical applications is evaluated, such as how to reduce power consumption or improve color accuracy while maintaining picture quality.
In a specific embodiment, the process of executing step 103 may specifically include the following steps:
(1) Performing backlight spectrum analysis on the backlight module model based on an initial picture optimization execution strategy to obtain spectrum distribution data;
(2) Constructing a linear programming model of the backlight module model, wherein the linear programming model aims at maximizing luminous efficiency and ensuring sufficient color gamut coverage;
(3) Carrying out luminous efficiency optimization parameter calculation on the light distribution data through a luminous efficiency optimization objective function in a linear programming model to obtain luminous efficiency optimization parameters, wherein the luminous efficiency optimization objective function is as follows: ，/> Representing luminous efficiency,/> Representing wavelength/>Spectral intensity under,/>Indicating the wavelength of the human eye/>Relative visual response under,/>Representing the total power consumption of the backlight module;
(4) Performing color gamut coverage optimization parameters on the light distribution data through color gamut coverage constraint conditions in the linear programming model, wherein the color gamut coverage constraint conditions are as follows: subjectto: ，/> Representing wavelength/> Spectral intensity under,/>Represents the/>Seed color at wavelength/>Spectral response function,/>Represents the/>Minimum coverage requirements for seed color;
(5) And generating initial backlight spectrum optimization parameters of the backlight module model according to the luminous efficiency optimization parameters and the color gamut coverage optimization parameters.
Specifically, backlight spectrum analysis is performed on the backlight module model based on an initial picture optimization execution strategy. And obtaining spectrum distribution data of the backlight module through optical simulation, namely, the intensity distribution of light under different wavelengths. For example, for a backlight module of a liquid crystal display, this involves analyzing the spectral output of different types of LEDs (e.g., blue LEDs and yellow LEDs) and how these light sources cooperate to produce the desired white light. And constructing a linear programming model of the backlight module model, maximizing luminous efficiency, ensuring sufficient color gamut coverage, and balancing power consumption (i.e. energy efficiency) and color performance. The luminous efficiency optimization objective function considers the spectral intensity of the backlight module and the relative visual response of human eyes to light with different wavelengths, and is compared with the total power consumption. For example, by optimizing the ratio of blue and yellow LEDs, the brightness of the display screen may be increased while reducing power consumption. In the linear planning model, the luminous efficiency optimization parameters are calculated. By applying the luminous efficiency optimization objective function, the ideal brightness level and color temperature of each LED are determined, so that the backlight module provides the optimal light output under the condition of lowest power consumption. For example, when designing a large LED screen for outdoor display, optimization parameters may tend to increase brightness and color saturation to accommodate viewing conditions in direct sunlight. And meanwhile, performing color gamut coverage optimization parameter calculation on the light distribution data through color gamut coverage constraint conditions in the linear programming model. Ensuring that the display screen is able to cover a sufficiently wide color gamut to exhibit rich and accurate colors. The distribution of the different colors over the entire visible spectrum is evaluated and it is ensured that each color meets at least its minimum coverage requirement. And generating initial backlight spectrum optimization parameters of the backlight module model according to the luminous efficiency optimization parameters and the color gamut coverage optimization parameters. These parameters directly affect the performance of the final display, including brightness, color reproduction, power consumption, and cost.
In a specific embodiment, the process of executing step 104 may specifically include the following steps:
(1) Carrying out population initialization on initial backlight spectrum optimization parameters through a preset MOEA/D-SFLA algorithm to generate a plurality of first backlight spectrum optimization parameters;
(2) Respectively calculating a first fitness value of each first backlight spectrum optimization parameter through a preset fitness function, and comparing the first fitness value with a first target value and a second target value, wherein the first target value is smaller than the second target value;
(3) If the first fitness value is less than the first target value, dividing the corresponding first backlight spectrum optimization parameter into a first spectrum optimization parameter group, if the first target value is less than the first fitness value is less than the second target value, dividing the corresponding first backlight spectrum optimization parameter into a second spectrum optimization parameter group, and if the second target value is less than the first fitness value, dividing the corresponding first backlight spectrum optimization parameter into a third spectrum optimization parameter group;
(4) Propagating, intersecting and mutating the first spectrum optimization parameter group and the second spectrum optimization parameter group, and intersecting and mutating the third spectrum optimization parameter group to obtain a plurality of second backlight spectrum optimization parameters;
(5) And respectively calculating second fitness values of each second backlight spectrum optimization parameter, and carrying out optimization analysis on a plurality of second backlight spectrum optimization parameters according to the second fitness values to obtain target backlight spectrum optimization parameters.
Specifically, the initial backlight spectrum optimization parameters are subjected to population initialization through a MOEA/D-SFLA algorithm, and a set of diversified candidate solutions, namely first backlight spectrum optimization parameters, are generated. For example, for a particular display module, these parameters include the intensity and color temperature distribution of the different LED light sources to meet specific display effect and power consumption requirements. And evaluating each first backlight spectrum optimization parameter through a preset fitness function, and calculating first fitness values of the first backlight spectrum optimization parameters. These fitness values reflect the performance of each parameter at a particular target, such as luminance uniformity, color reproduction capability, and energy efficiency. In calculating the fitness value, these values are compared with the set first and second target values. The two target values represent different performance criteria, wherein the first target value is smaller than the second target value. The first backlight spectrum optimization parameter is divided into different groups according to the comparison result of the first fitness value and the two target values. If the first fitness value is less than the first target value, the corresponding parameter is divided into a first population of spectrally optimized parameters; dividing into a second population if the first fitness value is between the two target values; if the target value is larger than the second target value, dividing into a third group. This classification mechanism ensures that parameters of different performance levels are reasonably classified and provides guidance for subsequent evolutionary operations. The operations of propagation, crossover and mutation are performed on the first and second populations of spectrally optimized parameters, while only crossover and mutation are performed on the third population. These operations simulate the process of natural selection and genetic variation to promote the evolution of solutions to a more optimal state. For example, in optimizing the spectral distribution of a display screen, the advantages of two different parameters can be combined by a crossover operation, or a new solution can be introduced by mutation to explore a wider solution space. And calculating a second fitness value for each second backlight spectrum optimization parameter, and performing optimization analysis based on the values. And screening the optimal backlight spectrum optimization parameters from all the candidate solutions. For example, optimization analysis may determine a specific set of LED light source distributions that not only provide high brightness and wide color gamut coverage, but also achieve energy efficiency.
In a specific embodiment, the process of executing step 105 may specifically include the following steps:
(1) Performing LED array analysis on the backlight module model based on an initial picture optimization execution strategy to obtain current LED array layout information;
(2) And performing light intensity distribution calculation on the backlight module model according to the current LED array layout information to obtain the light intensity distribution of the current LED array, wherein the light intensity distribution calculation function is as follows: ，/> representing the light intensity at coordinate point (x, y)/> Is a Fourier coefficient representing the contribution of different frequency components to the light intensity distribution, each/>Corresponding to preset frequency components, responsible for describing a specific pattern in the spatial light intensity distribution of the LED array,/>Frequency index along x and y directions,/>, respectivelyRepresenting the dimensions of the LED array in the x and y directions,/>Representing a spatially varying pattern in the light intensity distribution as a complex exponential function;
(3) Performing LED array optimization on the light intensity distribution of the current LED array by using a Fourier series method, and determining target LED array optimization parameters, wherein the target LED array optimization parameters comprise optimal LED layout and angular intensity distribution, and the Fourier series function is as follows: ，/> Representing the contribution of different frequency components to the light intensity distribution as Fourier coefficients,/> Representing the dimensions of the LED array in the x and y directions,Representing the light intensity at coordinate point (x, y)/>Representing the frequency index along the x and y directions,Representing an inverse form of a complex exponential function for transforming the light intensity distribution from the spatial domain to the frequency domain in a fourier transform process,/>To integrate the entire LED array area, it is shown that the contribution of the entire array to each fourier coefficient is a cumulative and integrated result.
Specifically, the LED array analysis is performed on the backlight module model based on the initial picture optimization execution strategy, and the layout information of the current LED array is obtained. Including the location, number, and brightness level of each LED. For example, the LED arrays are arranged in a uniform grid, and the brightness of each LED needs to be precisely controlled to ensure uniform brightness across the display screen. And calculating the light intensity distribution of the backlight module model according to the current LED array layout information. Fourier series analysis is applied and the spatial distribution of the light intensity is optimized. The light intensity distribution calculation function expresses the light intensity at a specific coordinate point (x, y) using a fourier series. This function takes into account the contributions of the individual LEDs to the spot light intensity, wherein the fourier coefficientsRepresenting the magnitude of the contribution of the different frequency components to the light intensity distribution, whereas/>This complex exponential function then expresses a spatially varying pattern in the light intensity distribution. And optimizing the light intensity distribution of the current LED array by using a Fourier series method. The fourier series function is used to calculate the contribution of the different frequency components to the light intensity distribution, thereby determining the target LED array optimization parameters. These parameters include optimal LED layout and angular intensity distribution, aimed at improving the light efficiency and visual quality of the whole display screen. For example, by optimizing the LED layout and light intensity distribution, it is possible to ensure that the monitor provides sharper and balanced brightness while displaying high contrast images, while reducing power consumption. The application of the fourier series function covers the calculation of the integral of the entire LED array area, taking into account the contribution of the entire array to each fourier coefficient. The integrated consideration enables the optimization process to comprehensively evaluate and improve the light intensity distribution. Through the above steps, the light intensity distribution is converted from the spatial domain to the frequency domain, so that the optimization process is more accurate and efficient. A group of target LED array optimization parameters are obtained, and the parameters not only improve the display effect, but also consider the energy efficiency and the cost.
In a specific embodiment, the process of executing step 106 may specifically include the following steps:
(1) Performing strategy optimization on the initial picture optimization execution strategy according to the target backlight spectrum optimization parameters and the target LED array optimization parameters to obtain a plurality of candidate picture optimization execution strategies;
(2) Respectively carrying out module integration test on each candidate picture optimization execution strategy through a backlight module model to obtain module integration test data of each candidate picture optimization execution strategy;
(3) Respectively calculating strategy evaluation indexes of each candidate picture optimization execution strategy according to the module integration test data;
(4) And carrying out optimization solution on the multiple candidate picture optimization execution strategies according to the strategy evaluation indexes to obtain a target picture optimization execution strategy.
Specifically, the initial picture optimization execution strategy is subjected to strategy optimization according to the target backlight spectrum optimization parameters and the target LED array optimization parameters, and a plurality of candidate picture optimization execution strategies are obtained. And adjusting and optimizing an initial picture display strategy according to key parameters such as spectral efficiency, color gamut coverage and the like. For example, if the target spectral optimization parameter points to a higher blue light output, the policy optimization involves adjusting the brightness of the blue LED or changing the configuration of the optical elements associated with the blue LED. And carrying out module integration test on each candidate picture optimization execution strategy through the backlight module model, and evaluating the performance of each strategy in actual application. In performing the module integration test, simulation software is used to reproduce the effects of different strategies on picture quality, such as brightness distribution, color uniformity, and overall visual effect. The module integration test data of each candidate frame optimization execution strategy, such as brightness level, color accuracy and energy consumption, are key indexes for evaluating the performance of the candidate frames. For example, one strategy increases power consumption while increasing brightness, while another strategy sacrifices some brightness while maintaining lower power consumption. And respectively calculating strategy evaluation indexes of each candidate picture optimization execution strategy according to the module integration test data. This includes quantifying the performance of each strategy, such as luminance uniformity, color accuracy, power consumption, etc., and comparing these metrics to target spectral optimization parameters. For example, one strategy may perform well in terms of color accuracy, but its power consumption exceeds a predetermined goal, while another strategy may provide acceptable color accuracy while maintaining lower power consumption. And carrying out optimization solution on the multiple candidate picture optimization execution strategies based on the strategy evaluation indexes, thereby obtaining the target picture optimization execution strategy. The advantages and disadvantages of each strategy are comprehensively considered, and the strategy which is best represented on a plurality of indexes is selected.
The method for optimizing a picture based on a backlight of a display module in the embodiment of the present application is described above, and the system for optimizing a picture based on a backlight of a display module in the embodiment of the present application is described below, referring to fig. 2, an embodiment of the system for optimizing a picture based on a backlight of a display module in the embodiment of the present application includes:
the creation module 201 is configured to create a backlight module model through LightTools software and module performance constraint conditions, and perform optical simulation and backlight module parameter acquisition on the backlight module model through MATLAB software to obtain a plurality of target backlight module parameters;
the definition module 202 is configured to define a state space and an action space of the backlight module model according to the plurality of target backlight module parameters, and perform a picture optimization execution policy analysis on the backlight module model through a dual-depth Q network algorithm to obtain an initial picture optimization execution policy;
the solving module 203 is configured to perform backlight spectrum analysis on the backlight module model based on the initial picture optimization execution policy to obtain spectrum distribution data, and perform linear programming solving on the spectrum distribution data to obtain initial backlight spectrum optimization parameters;
the processing module 204 is configured to perform local dimming optimization on the initial backlight spectrum optimization parameter through a preset MOEA/D-SFLA algorithm, so as to obtain a target backlight spectrum optimization parameter;
The analysis module 205 is configured to perform LED array analysis and fourier series optimization on the backlight module model based on the initial picture optimization execution policy, so as to obtain a target LED array optimization parameter;
And the optimization module 206 is configured to perform policy optimization and module integration test on the initial picture optimization execution policy according to the target backlight spectrum optimization parameter and the target LED array optimization parameter, so as to obtain a target picture optimization execution policy.
By combining the advanced optical simulation of the LightTools and MATLAB software with the Dual Depth Q Network (DDQN) algorithm and the MOEA/D-SFLA algorithm, not only is a single picture quality or energy efficiency focused, but also multi-dimensional comprehensive optimization of light intensity distribution, spectral distribution, energy consumption, cost and the like is realized. This allows the final display effect to reach an optimal balance in terms of brightness uniformity, color realism, energy efficiency, etc. By using the LightTools software to accurately model the backlight module and combining MATLAB to perform optical simulation and parameter acquisition, the propagation, scattering and reflection behaviors of light rays in the backlight module can be analyzed in detail, and accurate input data can be provided for subsequent optimization. This accurate modeling and analysis is the key basis for optimizing the effect. By applying an advanced double-depth Q network algorithm, an optimal action strategy is automatically learned and predicted, the brightness of the LED and the structure of an optical element can be intelligently adjusted, and different display requirements and use scenes can be dynamically adapted. This intelligent strategy greatly improves the efficiency and effectiveness of the optimization. By combining with the MOEA/D-SFLA algorithm, the problems of global optimization and local dimming optimization can be effectively solved, the local area can be finely adjusted while the optimal solution is ensured to be found in the global range, and the picture quality is further improved. Not only is the optimization performed at the component level, but also the integration and coordination of the entire backlight module system is concerned. Through simulation and integration test of a system level, the cooperative work of each component and strategy is ensured, and further the picture optimization accuracy of the backlight of the display module is improved.
It will be clear to those skilled in the art that, for convenience and brevity of description, specific working procedures of the above-described systems, apparatuses and units may refer to corresponding procedures in the foregoing method embodiments, which are not repeated herein.
The integrated units, if implemented in the form of software functional units and sold or used as stand-alone products, may be stored in a computer readable storage medium. Based on such understanding, the technical solution of the present application may be embodied essentially or in part or all of the technical solution or in part in the form of a software product stored in a storage medium, including instructions for causing a computer device (which may be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the method according to the embodiments of the present application. And the aforementioned storage medium includes: a U-disk, a removable hard disk, a read-only memory (ROM), a random access memory (random access memory, RAM), a magnetic disk, or an optical disk, or other various media capable of storing program codes.
The above embodiments are only for illustrating the technical solution of the present application, and not for limiting the same; although the application has been described in detail with reference to the foregoing embodiments, it will be understood by those of ordinary skill in the art that: the technical scheme described in the foregoing embodiments can be modified or some technical features thereof can be replaced by equivalents; such modifications and substitutions do not depart from the spirit and scope of the technical solutions of the embodiments of the present application.
Claims (6)
1. The picture optimization method based on the display module backlight is characterized by comprising the following steps of:
Creating a backlight module model through LightTools software and module performance constraint conditions, and performing optical simulation and backlight module parameter acquisition on the backlight module model through MATLAB software to obtain a plurality of target backlight module parameters;
Defining a state space and an action space of the backlight module model according to the target backlight module parameters, and performing picture optimization execution strategy analysis on the backlight module model through a double-depth Q network algorithm to obtain an initial picture optimization execution strategy; the method specifically comprises the following steps: defining a state space of the backlight module model according to the target backlight module parameters, wherein the state space is as follows: a light intensity distribution factor and a spectral distribution factor; defining an action space of the backlight module model according to the target backlight module parameters, wherein the action space comprises: brightness adjustment of the LEDs and structural adjustment of the optical elements; inputting a plurality of target backlight module parameters of the backlight module model into a double-depth Q network algorithm, wherein the double-depth Q network algorithm comprises a first depth neural network and a second depth neural network; the Q value of the current state of the backlight module model is estimated through the first deep neural network to obtain the Q value of the current state, and the Q value of the current state of the backlight module model is updated through the second deep neural network to obtain the Q value of the next state, wherein the Q value updating formula is as follows: ， Expressed in state/> Down execution action/>When according to policy/>Expected benefit obtained,/>For learning rate, determining the degree to which the newly received information covers the old information,/>Representing an instant prize, reflecting the immediate effect of the current action,Representing discount factors, for measuring importance of future rewards,/>Representing the current state,/>The next state is indicated and the next state is indicated,Representing the current action,/>Representing the next action,/>Representing network parameters,/>Is the current value,/>Is old value/>Expressed in the next state/>Optional all actions/>Selected to make/>The maximum Q value of the motion; and calculating a target Q value of the backlight module model according to the Q value of the next state to obtain the target Q value, wherein the target Q value calculation function is as follows: /(I)，/>Represents the/>Target Q value of individual state-action pair,/>Represents the/>Instant rewards earned by individual state-action pairs,/>Representing a discount factor, representing the present value of a future reward,/>Indicating the next state to transition to after the action is performed,Representing network parameters,/>Is the current value,/>Is old value/>Representing an old version of the network parameters,Representation selection enables/>Maximized action/>; Performing bonus function calculation on the target Q value to obtain a bonus function calculation result, and performing picture optimization execution strategy analysis on the backlight module model according to the bonus function calculation result to obtain an initial picture optimization execution strategy;
performing backlight spectrum analysis on the backlight module model based on the initial picture optimization execution strategy to obtain spectrum distribution data, and performing linear programming solution on the spectrum distribution data to obtain initial backlight spectrum optimization parameters;
Carrying out local dimming optimization on the initial backlight spectrum optimization parameters through a preset MOEA/D-SFLA algorithm to obtain target backlight spectrum optimization parameters; the method specifically comprises the following steps: carrying out population initialization on the initial backlight spectrum optimization parameters through a preset MOEA/D-SFLA algorithm to generate a plurality of first backlight spectrum optimization parameters; respectively calculating a first fitness value of each first backlight spectrum optimization parameter through a preset fitness function, and comparing the first fitness value with a first target value and a second target value, wherein the first target value is smaller than the second target value; if the first fitness value is less than the first target value, dividing the corresponding first backlight spectrum optimization parameter into a first spectrum optimization parameter group, if the first target value is less than the first fitness value is less than the second target value, dividing the corresponding first backlight spectrum optimization parameter into a second spectrum optimization parameter group, and if the second target value is less than the first fitness value, dividing the corresponding first backlight spectrum optimization parameter into a third spectrum optimization parameter group; propagating, intersecting and mutating the first spectrum optimization parameter group and the second spectrum optimization parameter group, and intersecting and mutating the third spectrum optimization parameter group to obtain a plurality of second backlight spectrum optimization parameters; respectively calculating second fitness values of each second backlight spectrum optimization parameter, and carrying out optimization analysis on the plurality of second backlight spectrum optimization parameters according to the second fitness values to obtain target backlight spectrum optimization parameters;
Performing LED array analysis and Fourier series optimization on the backlight module model based on the initial picture optimization execution strategy to obtain target LED array optimization parameters;
And carrying out strategy optimization and module integration test on the initial picture optimization execution strategy according to the target backlight spectrum optimization parameters and the target LED array optimization parameters to obtain a target picture optimization execution strategy.
2. The method for optimizing a picture based on a backlight of a display module according to claim 1, wherein the creating a backlight module model by LightTools software and module performance constraint conditions, and performing optical simulation and backlight module parameter acquisition on the backlight module model by MATLAB software, obtaining a plurality of target backlight module parameters, comprises:
Obtaining a module performance constraint condition, wherein the module performance constraint condition comprises: power consumption limitations, cost budgets, and size limitations;
creating a backlight module model according to the module performance constraint conditions by using LightTools software, wherein the backlight module model comprises N backlight modules;
optical simulation is carried out on the backlight module model through MATLAB software, and light intensity calculation is carried out on the backlight module model to obtain light intensity, wherein the light intensity calculation function is as follows: ，/> representing coordinates in three-dimensional space Light intensity at,/>Represents the/>Initial light intensity of individual LEDs,/>Represents the/>Light attenuation coefficient of individual LEDs,/>The representation is from the/>Individual LED to position/>Distance of/>From the/>Individual LED to position/>The included angle between the light rays and the normal line, N represents the total number of LEDs;
And performing backlight module power consumption calculation on the backlight module model to obtain the total power consumption of the backlight module, wherein the backlight module power consumption calculation function is as follows: ，/> Representing the total power consumption of the backlight module First/>Operating voltage of individual LEDs,/>Represents the/>Operating current of individual LEDs,/>Represents the/>The ageing coefficient of the individual LEDs is such that,Indicating the operating time of the LED;
And calculating the spectral efficiency of the backlight module model to obtain the spectral efficiency of the backlight module, wherein the spectral efficiency calculation function is as follows: ，/> the spectral efficiency of the backlight module is expressed as the light output effect per unit power consumption,/> Expressed in wavelength/>Spectral distribution function at,/>Representing the corresponding wavelength of the human eye/>Describing the sensitivity of the human eye to light of different wavelengths,/>Representing the minimum value of the wavelength range,/>Represents the maximum of the wavelength range;
performing color gamut coverage calculation on the backlight module model to obtain color gamut coverage, wherein the color gamut coverage calculation function is as follows: ，/> For color gamut coverage, represent the ability of the backlight module to cover the target color gamut,/> Color/>At wavelength/>Spectral distribution function at,/>Representing color/>Representing the importance of color c in the overall gamut,/>Representing color/>Is used for the target coverage of the (c) in the (c),Representing a gamut coverage function describing contributions of different wavelengths of light to the gamut coverage;
And generating a plurality of target backlight module parameters of the backlight module model according to the light intensity, the total power consumption of the backlight module, the spectral efficiency of the backlight module and the color gamut coverage.
3. The method for optimizing a display screen based on a backlight module according to claim 1, wherein the performing backlight spectrum analysis on the backlight module model based on the initial image optimization execution strategy to obtain spectrum distribution data, and performing linear programming solution on the spectrum distribution data to obtain initial backlight spectrum optimization parameters comprises:
Performing backlight spectrum analysis on the backlight module model based on the initial picture optimization execution strategy to obtain spectrum distribution data;
constructing a linear programming model of the backlight module model, wherein the linear programming model aims at maximizing luminous efficiency and ensuring sufficient color gamut coverage;
And calculating the luminous efficiency optimization parameters of the spectrum distribution data through a luminous efficiency optimization objective function in the linear programming model to obtain the luminous efficiency optimization parameters, wherein the luminous efficiency optimization objective function is as follows: ，/> Representing luminous efficiency,/> Representing wavelength/>The intensity of the spectrum of light at the bottom,Indicating the wavelength of the human eye/>Relative visual response under,/>Representing the total power consumption of the backlight module;
And performing color gamut coverage optimization parameters on the spectrum distribution data through color gamut coverage constraint conditions in the linear programming model, wherein the color gamut coverage constraint conditions are as follows: ， Representing wavelength/> Spectral distribution function,/>Represents the/>Seed color at wavelength/>Spectral response function,/>Represents the/>Minimum coverage requirements for seed color;
And generating initial backlight spectrum optimization parameters of the backlight module model according to the luminous efficiency optimization parameters and the color gamut coverage optimization parameters.
4. The method for optimizing a backlight frame based on a display module according to claim 1, wherein the performing LED array analysis and fourier series optimization on the backlight module model based on the initial frame optimization execution policy to obtain a target LED array optimization parameter comprises:
Performing LED array analysis on the backlight module model based on the initial picture optimization execution strategy to obtain current LED array layout information;
And performing light intensity distribution calculation on the backlight module model according to the current LED array layout information to obtain the light intensity distribution of the current LED array, wherein the light intensity distribution calculation function is as follows: ，/> representing the light intensity at coordinate point (x, y)/> Is a Fourier coefficient representing the contribution of different frequency components to the light intensity distribution, each/>Corresponding to preset frequency components, responsible for describing a specific pattern in the spatial light intensity distribution of the LED array,/>Frequency index along x and y directions,/>, respectivelyRepresenting the dimensions of the LED array in the x and y directions,/>As a complex exponential function, representing a spatially varying pattern in the light intensity distribution, N represents a range of frequency indices along the x-direction, M represents a range of frequency indices along the y-direction;
Performing LED array optimization on the light intensity distribution of the current LED array by using a Fourier series method, and determining target LED array optimization parameters, wherein the target LED array optimization parameters comprise optimal LED layout and angular intensity distribution, and the Fourier series function is as follows: ，/> Representing the contribution of different frequency components to the light intensity distribution as Fourier coefficients,/> Representing the dimensions of the LED array in the x and y directions,/>Representing the light intensity at coordinate point (x, y)/>Representing the frequency index along the x and y directions,Representing an inverse form of a complex exponential function for transforming the light intensity distribution from the spatial domain to the frequency domain in a fourier transform process,/>To integrate the entire LED array area, it is shown that the contribution of the entire array to each fourier coefficient is a cumulative and integrated result.
5. The method for optimizing a display module backlight-based frame according to claim 4, wherein the performing a policy optimization and a module integration test on the initial frame optimization execution policy according to the target backlight spectrum optimization parameter and the target LED array optimization parameter to obtain a target frame optimization execution policy comprises:
performing strategy optimization on the initial picture optimization execution strategy according to the target backlight spectrum optimization parameters and the target LED array optimization parameters to obtain a plurality of candidate picture optimization execution strategies;
performing module integration test on each candidate picture optimization execution strategy through the backlight module model to obtain module integration test data of each candidate picture optimization execution strategy;
Respectively calculating strategy evaluation indexes of each candidate picture optimization execution strategy according to the module integration test data;
and carrying out optimization solution on the candidate picture optimization execution strategies according to the strategy evaluation indexes to obtain target picture optimization execution strategies.
6. The utility model provides a picture optimizing system based on display module assembly backlight which characterized in that, the picture optimizing system based on display module assembly backlight includes:
The system comprises a creation module, a module selection module and a module performance constraint condition module, wherein the creation module is used for creating a backlight module model through the LightTools software and the module performance constraint condition, and carrying out optical simulation and backlight module parameter acquisition on the backlight module model through MATLAB software to obtain a plurality of target backlight module parameters;
The definition module is used for defining a state space and an action space of the backlight module model according to the target backlight module parameters, and carrying out picture optimization execution strategy analysis on the backlight module model through a double-depth Q network algorithm to obtain an initial picture optimization execution strategy; the method specifically comprises the following steps: defining a state space of the backlight module model according to the target backlight module parameters, wherein the state space is as follows: a light intensity distribution factor and a spectral distribution factor; defining an action space of the backlight module model according to the target backlight module parameters, wherein the action space comprises: brightness adjustment of the LEDs and structural adjustment of the optical elements; inputting a plurality of target backlight module parameters of the backlight module model into a double-depth Q network algorithm, wherein the double-depth Q network algorithm comprises a first depth neural network and a second depth neural network; the Q value of the current state of the backlight module model is estimated through the first deep neural network to obtain the Q value of the current state, and the Q value of the current state of the backlight module model is updated through the second deep neural network to obtain the Q value of the next state, wherein the Q value updating formula is as follows: ， Expressed in state/> Down execution action/>When according to policy/>Expected benefit obtained,/>For learning rate, determining the degree to which the newly received information covers the old information,/>Representing instant rewards, reflecting the direct effect of the current action,/>Representing discount factors, for measuring importance of future rewards,/>Representing the current state,/>Representing the next state,/>Representing the current action,/>Representing the next action,/>Representing network parameters,/>Is the current value,/>Is old value/>Expressed in the next state/>Optional all actions/>Selected to make/>The maximum Q value of the motion; and calculating a target Q value of the backlight module model according to the Q value of the next state to obtain the target Q value, wherein the target Q value calculation function is as follows: /(I)，/>Represents the/>Target Q value of individual state-action pair,/>Represents the/>Instant rewards earned by individual state-action pairs,/>Representing a discount factor, representing the present value of a future reward,/>Indicating the next state to transition to after the action is performed,Representing network parameters,/>Is the current value,/>Is old value/>Representing an old version of the network parameters,Representation selection enables/>Maximized action/>; Performing bonus function calculation on the target Q value to obtain a bonus function calculation result, and performing picture optimization execution strategy analysis on the backlight module model according to the bonus function calculation result to obtain an initial picture optimization execution strategy;
The solving module is used for carrying out backlight spectrum analysis on the backlight module model based on the initial picture optimization execution strategy to obtain spectrum distribution data, and carrying out linear programming solving on the spectrum distribution data to obtain initial backlight spectrum optimization parameters;
The processing module is used for carrying out local dimming optimization on the initial backlight spectrum optimization parameters through a preset MOEA/D-SFLA algorithm to obtain target backlight spectrum optimization parameters; the method specifically comprises the following steps: carrying out population initialization on the initial backlight spectrum optimization parameters through a preset MOEA/D-SFLA algorithm to generate a plurality of first backlight spectrum optimization parameters; respectively calculating a first fitness value of each first backlight spectrum optimization parameter through a preset fitness function, and comparing the first fitness value with a first target value and a second target value, wherein the first target value is smaller than the second target value; if the first fitness value is less than the first target value, dividing the corresponding first backlight spectrum optimization parameter into a first spectrum optimization parameter group, if the first target value is less than the first fitness value is less than the second target value, dividing the corresponding first backlight spectrum optimization parameter into a second spectrum optimization parameter group, and if the second target value is less than the first fitness value, dividing the corresponding first backlight spectrum optimization parameter into a third spectrum optimization parameter group; propagating, intersecting and mutating the first spectrum optimization parameter group and the second spectrum optimization parameter group, and intersecting and mutating the third spectrum optimization parameter group to obtain a plurality of second backlight spectrum optimization parameters; respectively calculating second fitness values of each second backlight spectrum optimization parameter, and carrying out optimization analysis on the plurality of second backlight spectrum optimization parameters according to the second fitness values to obtain target backlight spectrum optimization parameters;
the analysis module is used for carrying out LED array analysis and Fourier series optimization on the backlight module model based on the initial picture optimization execution strategy to obtain target LED array optimization parameters;
And the optimization module is used for carrying out strategy optimization and module integration test on the initial picture optimization execution strategy according to the target backlight spectrum optimization parameter and the target LED array optimization parameter to obtain a target picture optimization execution strategy.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410155676.1A CN117690388B (en) | 2024-02-04 | 2024-02-04 | Picture optimization method and system based on display module backlight |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410155676.1A CN117690388B (en) | 2024-02-04 | 2024-02-04 | Picture optimization method and system based on display module backlight |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN117690388A CN117690388A (en) | 2024-03-12 |
| CN117690388B true CN117690388B (en) | 2024-04-19 |
Family
ID=90133867
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202410155676.1A Active CN117690388B (en) | 2024-02-04 | 2024-02-04 | Picture optimization method and system based on display module backlight |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN117690388B (en) |
Citations (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107342056A (en) * | 2017-07-31 | 2017-11-10 | 天津大学 | A kind of region backlight dynamic light adjustment method that the algorithm that leapfrogs is shuffled based on improvement |
| KR20200017598A (en) * | 2018-08-01 | 2020-02-19 | 가톨릭대학교 산학협력단 | Method and device for classifying data using dual machine learning model |
| CN113674705A (en) * | 2021-08-27 | 2021-11-19 | 天津大学 | Backlight extraction method based on radial basis function neural network proxy model assisted particle swarm algorithm |
| WO2021235653A1 (en) * | 2020-05-19 | 2021-11-25 | 삼성전자주식회사 | Method and display device for modifying algorithm for calculating gradation filter |
| CN114880914A (en) * | 2021-08-27 | 2022-08-09 | 天津大学 | Regional dimming model based on multi-objective optimization |
| CN115660992A (en) * | 2022-11-01 | 2023-01-31 | 南京宁麒智能计算芯片研究院有限公司 | Local backlight dimming method, system, device and medium |
| KR20230027688A (en) * | 2021-08-19 | 2023-02-28 | 삼성전자주식회사 | Method and electronic device for optimizing image quality |
| KR20230029412A (en) * | 2021-08-24 | 2023-03-03 | 삼성전자주식회사 | Display device and operating method thereof |
| CN116451577A (en) * | 2023-04-11 | 2023-07-18 | 中国矿业大学 | Product change path multi-target optimization method integrating reinforcement learning and differential evolution |
| CN117348831A (en) * | 2023-12-06 | 2024-01-05 | 深圳康荣电子有限公司 | Picture adjustment method and system for liquid crystal display screen |
| WO2024014706A1 (en) * | 2022-07-13 | 2024-01-18 | 삼성전자주식회사 | Electronic device for training neural network model performing image enhancement, and control method therefor |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US11120303B2 (en) * | 2018-12-17 | 2021-09-14 | King Fahd University Of Petroleum And Minerals | Enhanced deep reinforcement learning deep q-network models |
-
2024
- 2024-02-04 CN CN202410155676.1A patent/CN117690388B/en active Active
Patent Citations (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107342056A (en) * | 2017-07-31 | 2017-11-10 | 天津大学 | A kind of region backlight dynamic light adjustment method that the algorithm that leapfrogs is shuffled based on improvement |
| KR20200017598A (en) * | 2018-08-01 | 2020-02-19 | 가톨릭대학교 산학협력단 | Method and device for classifying data using dual machine learning model |
| WO2021235653A1 (en) * | 2020-05-19 | 2021-11-25 | 삼성전자주식회사 | Method and display device for modifying algorithm for calculating gradation filter |
| KR20230027688A (en) * | 2021-08-19 | 2023-02-28 | 삼성전자주식회사 | Method and electronic device for optimizing image quality |
| KR20230029412A (en) * | 2021-08-24 | 2023-03-03 | 삼성전자주식회사 | Display device and operating method thereof |
| CN113674705A (en) * | 2021-08-27 | 2021-11-19 | 天津大学 | Backlight extraction method based on radial basis function neural network proxy model assisted particle swarm algorithm |
| CN114880914A (en) * | 2021-08-27 | 2022-08-09 | 天津大学 | Regional dimming model based on multi-objective optimization |
| WO2024014706A1 (en) * | 2022-07-13 | 2024-01-18 | 삼성전자주식회사 | Electronic device for training neural network model performing image enhancement, and control method therefor |
| CN115660992A (en) * | 2022-11-01 | 2023-01-31 | 南京宁麒智能计算芯片研究院有限公司 | Local backlight dimming method, system, device and medium |
| CN116451577A (en) * | 2023-04-11 | 2023-07-18 | 中国矿业大学 | Product change path multi-target optimization method integrating reinforcement learning and differential evolution |
| CN117348831A (en) * | 2023-12-06 | 2024-01-05 | 深圳康荣电子有限公司 | Picture adjustment method and system for liquid crystal display screen |
Also Published As
| Publication number | Publication date |
|---|---|
| CN117690388A (en) | 2024-03-12 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Liu et al. | A future land use simulation model (FLUS) for simulating multiple land use scenarios by coupling human and natural effects | |
| CN106296669B (en) | A kind of image quality evaluating method and device | |
| Nault et al. | Development and test application of the UrbanSOLve decision-support prototype for early-stage neighborhood design | |
| Dong et al. | Intelligent optimization: A novel framework to automatize multi-objective optimization of building daylighting and energy performances | |
| Luo et al. | Optimal product positioning with consideration of negative utility effect on consumer choice rule | |
| Newland et al. | Empirically derived method and software for semi-automatic calibration of Cellular Automata land-use models | |
| Krebs | An empirically grounded model of green electricity adoption in Germany: Calibration, validation and insights into patterns of diffusion | |
| CN109615414A (en) | House property predictor method, device and storage medium | |
| CN115983438A (en) | Method and device for determining operation strategy of data center terminal air conditioning system | |
| CN117348831B (en) | Picture adjustment method and system for liquid crystal display screen | |
| CN117690388B (en) | Picture optimization method and system based on display module backlight | |
| JP4320361B1 (en) | Future financial forecasting system, future financial forecasting method, and future financial forecasting program | |
| CN115482227B (en) | Machine vision self-adaptive imaging environment adjusting method | |
| CN112307537B (en) | Natural light vision and non-vision effect coupling optimization method of multi-objective optimization algorithm | |
| Li et al. | Co-learning patterns as emergent market phenomena: An electricity market illustration | |
| Moya et al. | Coral reefs optimization algorithms for agent-based model calibration | |
| Huang et al. | Online crowd learning through strategic worker reports | |
| Gomes et al. | Design of an adaptive mutation operator in an electrical load management case study | |
| Lau et al. | A reinforcement learning algorithm developed to model GenCo strategic bidding behavior in multidimensional and continuous state and action spaces | |
| Hong et al. | Fusion Evaluation of College Cultivation by Adaptive Multivariate Neural Network Model | |
| Abo et al. | Sampling density and frequency as data quality determinants in smart grids | |
| CN114330147B (en) | Model training method, color formula prediction method, system, device and medium | |
| Sakal et al. | Towards a Many-Objective Optimiser for University Course Timetabling | |
| CN116091403B (en) | Subjective and objective evaluation method for color night vision fusion image quality | |
| WO2023181320A1 (en) | Model processing device, model processing method, and recording medium |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |