A method and system for optimizing the dynamic contrast ratio of a liquid crystal display screen

Abstract

This application relates to the field of liquid crystal displays, and discloses a method and system for optimizing the dynamic contrast of a liquid crystal display screen, including the following steps: By collecting the ambient light intensity and backlight module temperature parameters in real time, and combining with the analysis of the image content entropy value, a non-uniform backlight partition grid is dynamically generated; Based on hardware parameters such as the refractive index of the light guide plate and the LED spacing, a physical model of brightness diffusion is constructed, and an improved particle swarm algorithm with temperature adaptability is used to solve the optimal brightness of each partition; The driving voltage is generated through piecewise non-linear gamma mapping and inverse function iterative calculation to adapt to the non-linear response characteristics of the backlight module; The diffusion equation and optimization calculation are executed in parallel by a hardware accelerator, and the model parameters are dynamically adjusted in combination with the feedback of the optical sensor to form a closed-loop control link. Through the dynamic partitioning adapted to the environment, the optimization algorithm driven by the physical model, and the real-time processing accelerated by the hardware, the present invention comprehensively improves the contrast performance and dynamic scene adaptability of the display screen.

Description

Technical Field

[0001] The present invention relates to the technical field of liquid crystal display, and particularly to a method and system for optimizing the dynamic contrast of a liquid crystal display screen. Background Art

[0002] In the existing liquid crystal display technology, the static partitioning method of the backlight partition is difficult to adapt to dynamic scenes and environmental changes. The fixed partition grid is prone to halo effects and brightness distortion in detail areas under complex lighting conditions. Especially when displaying high-dynamic-range content, it is difficult to balance the coordinated control of local highlights and dark-field details.

[0003] Traditional optimization algorithms mostly construct objective functions based on idealized assumptions, without fully considering the actual effects of hardware parameters such as the refractive characteristics of the light guide plate and the LED layout on brightness diffusion, resulting in a deviation between the algorithm output and the physical system response, and affecting the contrast optimization effect.

[0004] In addition, the global gamma correction strategy cannot adapt to the local brightness characteristics of non-uniform backlight partitions, is prone to color distortion in low gray-scale areas, and lacks a closed-loop feedback mechanism, making it difficult to cope with dynamic interferences such as backlight module aging and environmental mutations, resulting in a decline in display performance during long-term use.

[0005] The existing technologies have significant limitations in aspects such as the real-time performance of multi-objective optimization, the coupling degree of hardware parameters, and the environmental adaptability, restricting the realization of high-dynamic-contrast display effects. Summary of the Invention

[0006] Aiming at the deficiencies of the existing technologies, the present invention provides a method and system for optimizing the dynamic contrast of a liquid crystal display screen, solving the technical problems of dynamic scene contrast distortion, significant halo effects, poor environmental adaptability, and long-term display performance decline of existing liquid crystal display screens due to fixed backlight partitions, global gamma correction, and lack of closed-loop feedback.

[0007] To achieve the above objectives, the present invention is realized through the following technical solutions: A method for optimizing the dynamic contrast of a liquid crystal display screen includes the following steps:

[0008] Collect the brightness component of the input image and environmental parameters;

[0009] Dynamically generate a non-uniform backlight partition grid based on the image content;

[0010] Establish a collaborative optimization model for backlight brightness and solve the optimal brightness value for each partition;

[0011] Generate a non-linear gamma mapping curve according to the optimal brightness value;

[0012] Synthesize the drive signal and perform closed-loop feedback adjustment.

[0013] Preferably, the step of dynamically generating a non-uniform backlight zoning grid based on image content includes:

[0014] Traverse and calculate the local entropy value of each area of the image through a sliding window, and divide the image into multiple initial partitions according to a preset entropy value threshold, where the high entropy value area corresponds to the area with rich details;

[0015] Dynamically adjust the entropy value threshold according to the ambient light intensity, and merge adjacent partitions with an area smaller than the preset threshold to generate the final non-uniform backlight zoning grid.

[0016] Preferably, the specific rules for merging adjacent partitions with an area smaller than the preset threshold include:

[0017] Judge the difference in the average brightness of adjacent partitions. If the difference is less than the preset tolerance range, merge them;

[0018] The area of the merged partition does not exceed the size of the minimum controllable area of the backlight module.

[0019] Preferably, the step of establishing a backlight brightness collaborative optimization model and solving the optimal brightness value of each partition includes:

[0020] Construct a brightness diffusion model based on the refractive index of the light guide plate and the LED spacing parameters of the backlight module to simulate the brightness diffusion relationship between adjacent partitions;

[0021] Construct a multi-objective optimization model, which takes maximizing contrast, minimizing power consumption, and minimizing the liquid crystal response time as optimization objectives, and introduces a constraint condition on the brightness difference between adjacent partitions.

[0022] Preferably, the solution method of the multi-objective optimization model includes:

[0023] Use an improved particle swarm algorithm for iterative calculation, and the inertia weight of the algorithm is dynamically adjusted according to the real-time temperature of the backlight module;

[0024] During the particle update process, constrain the brightness difference between adjacent partitions not to exceed the preset threshold.

[0025] Preferably, the step of generating a non-linear gamma mapping curve according to the optimal brightness value includes:

[0026] Generate a non-linear gamma mapping curve for each partition according to the optimal brightness value of each partition, and the gamma mapping curve is obtained by fitting the error function that minimizes the target brightness and the actual brightness;

[0027] Based on the proportional relationship between the input brightness and the target brightness, map the input brightness to a drive voltage signal.

[0028] Preferably, the generation method of the non-linear gamma mapping curve includes:

[0029] Use a piecewise quadratic function to locally fit the gamma curve of each partition, where the piecewise interval is dynamically divided according to the optimal brightness value of the partition;

[0030] Convert the input brightness value into a driving voltage through inverse function calculation.

[0031] Preferably, the step of synthesizing the driving signal and performing closed-loop feedback adjustment includes:

[0032] Parallelly execute the matrix operation of the brightness diffusion model and the iterative calculation of the optimization algorithm through a hardware accelerator;

[0033] Dynamically adjust the diffusion coefficient and optimization weight parameters of the brightness diffusion model according to the actual brightness data fed back by the optical sensor.

[0034] Preferably, the hardware accelerator is implemented by FPGA or ASIC, and its architecture includes:

[0035] A parallel computing unit for block processing of the brightness diffusion equation of the backlight partition;

[0036] A pipeline scheduling module that allocates computing resources to partitions with high real-time requirements according to priorities.

[0037] The present invention also provides a dynamic contrast optimization system for a liquid crystal display screen, including:

[0038] A data acquisition module for acquiring image brightness components and environmental parameters;

[0039] A dynamic partitioning module for generating a non-uniform backlight partition grid based on the image entropy value;

[0040] A collaborative optimization module for establishing a brightness diffusion model and solving a multi-objective optimization problem;

[0041] A gamma mapping module for generating a non-linear gamma curve and synthesizing a driving signal;

[0042] An execution feedback module for executing driving through a hardware accelerator and performing closed-loop parameter adjustment.

[0043] The present invention provides a dynamic contrast optimization method and system for a liquid crystal display screen. It has the following beneficial effects:

[0044] 1. By real-time collecting the ambient light intensity and temperature parameters and dynamically adjusting the entropy value partition threshold and merging rule, the present invention can realize the intelligent adjustment of the spatial resolution of the backlight partition with the change of ambient light. It can suppress invalid sub-partitions in strong light environments and retain the independent control ability of detail areas in weak light environments, effectively avoiding the halo effect and detail loss caused by fixed partitions.

[0045] 2. The present invention constructs a brightness diffusion equation by combining hardware parameters such as the refractive index of the light guide plate and the LED pitch of the backlight module, embeds physical characteristics into the optimization objective function, and dynamically adjusts the algorithm weights through a temperature adaptive mechanism, which can suppress the sharp increase in driving power consumption in a high-temperature environment while ensuring high contrast, and achieve a balance between display effects and hardware security.

[0046] 3. The present invention generates a locally fitted gamma curve based on the optimal brightness value of each partition, and uses a piecewise quadratic function and Newton iterative inverse operation, which can overcome the modeling error of the traditional global gamma curve for non-linear response, achieve high-precision conversion from input brightness to driving voltage, and significantly reduce color distortion in the low gray-scale region.

[0047] 4. The present invention realizes parallel computing of the brightness diffusion equation and the optimization algorithm through the FPGA / ASIC architecture, combined with a dynamic priority scheduling strategy, which can meet the real-time processing requirements of high refresh rate scenarios (such as 120Hz game screens), and ensure the frame rate stability of large-size display screens under multi-partition control.

[0048] 5. The present invention corrects the diffusion coefficient and optimization weights in real time based on the optical sensor data, forms a closed-loop control link from parameter acquisition, model calculation to driving output, which can compensate for brightness deviations caused by factors such as backlight module aging and environmental mutations, and improve the consistency of the display system during long-term use. BRIEF DESCRIPTION OF THE DRAWINGS

[0049] Figure 1 is a schematic flowchart of the method of the present invention;

[0050] Figure 2 is a schematic structural diagram of the system of the present invention.

[0051] Among them, 10. Data acquisition module; 20. Dynamic partition module; 30. Cooperative optimization module; 40. Gamma mapping module; 50. Execution feedback module. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0052] The following will clearly and completely describe the technical solutions in the embodiments of the present invention with reference to the accompanying drawings of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0053] Please refer to the appended Figure 1 , the present invention provides a method for optimizing the dynamic contrast of a liquid crystal display screen, which improves the contrast performance through dynamic partition cooperative optimization and non-linear mapping technology.

[0054] Such as Figure 1As shown, the method for optimizing the dynamic contrast of the liquid crystal display screen may include the following steps:

[0055] S1. Collect the luminance component of the input image and the environmental parameters;

[0056] S2. Dynamically generate a non-uniform backlight partition grid based on the image content;

[0057] S3. Establish a collaborative optimization model for backlight luminance, and solve the optimal luminance value for each partition;

[0058] S4. Generate a non-linear gamma mapping curve according to the optimal luminance value;

[0059] S5. Synthesize the driving signal and perform closed-loop feedback adjustment.

[0060] The following is a detailed description of each step in the method of the present invention, and a comprehensive elaboration is carried out on the specific implementation principle, technical details and process of each step.

[0061] For step S1, its specific implementation process is as follows:

[0062] First, establish a calculation model for the luminance component of the input image. The luminance component is obtained by conversion through the CIE 1931 standard colorimetry formula, and the specific calculation relationship is:

[0063]

[0064] In the formula, , , respectively represent the red, green, and blue channel values of the input image at the pixel coordinates , and is the converted luminance component. This luminance component is used to characterize the light and dark distribution characteristics of each region in the image and provides a data basis for subsequent dynamic partitioning.

[0065] Subsequently, collect the ambient light intensity parameter through a light sensor. Exemplarily, the light sensor is integrated at the display screen border, and its spectral response range covers the visible light band. The measured value of the ambient light intensity is transmitted to the main control unit through the analog-to-digital conversion module for dynamically adjusting the sensitivity threshold of subsequent partitions.

[0066] Furthermore, obtain the real-time temperature parameter of the backlight module through a temperature sensor. The temperature sensor is mounted on the heat-sensitive area of the backlight driving circuit. The temperature parameter is dynamically bound to the weight coefficient in the optimization algorithm, and the specific binding relationship is:

[0067]

[0068] In the formula, is the reference weight coefficient, is the reference temperature value. This relationship can achieve real-time adaptation of algorithm parameters to the hardware state.

[0069] The said luminance component , ambient light intensity and temperature parameter are transmitted to the dynamic partitioning module through the data bus, constituting the input conditions for the subsequent steps.

[0070] Through the above steps, the standardized extraction of image luminance information can be achieved, and physical constraint conditions for dynamic optimization can be provided through multi-source environmental parameter acquisition to ensure the environmental adaptability of subsequent partitioning.

[0071] For step S2, its specific implementation process is as follows:

[0072] First, a dynamic partitioning generation model is established, and the local entropy value is calculated based on the input luminance component The specific formula is:

[0073]

[0074] In the formula, represents the occurrence probability of the gray level within the sliding window (exemplarily an 8×8 pixel window), is the coordinate and

[0075] is the local entropy value at the sub-region. The high-entropy value region corresponds to the detailed region with dense texture or edges in the image, and the low-entropy value region corresponds to the smooth or uniform region. This calculation can quantify the detail complexity of the image region and provide a basis for partitioning.

[0076]

[0077] In the formula, is the reference entropy value threshold (exemplarily taking a value of 4.0), is the reference ambient light intensity (exemplarily taking a value of 100 lux). When the ambient light intensifies, the threshold is increased proportionally to reduce the over-fine sub-region partitioning caused by external light interference.

[0078] Furthermore, a merging operation is performed on the initial partitions, and the merging rules are subject to the following constraint conditions:

[0079] 1. Luminance mean difference constraint: If the average luminance difference between adjacent partitions satisfies

[0080]

[0081] Then the merging is triggered. In the formula, , are the average brightness values of adjacent partitions, is exemplarily set to 10% of the brightness dynamic range (0 - 255), that is, 25.5.

[0082] 2. Minimum controllable area constraint: The area of the merged partition needs to satisfy

[0083]

[0084] In the formula is the size of the minimum controllable area of the backlight module, and the exemplary value is 16×16 pixels, which is determined by the physical addressing ability of the backlight driving chip (such as the resolution of the LED control unit).

[0085] After the dynamic partition grid is generated, it serves as the spatial constraint basis for subsequent brightness collaborative optimization.

[0086] For step S3, its specific implementation process is as follows:

[0087] First, construct a physical model for backlight brightness collaborative optimization, and establish a brightness diffusion equation based on the hardware parameters of the backlight module (refractive index of the light guide plate , LED pitch and light source wavelength ) to quantify the brightness coupling relationship between adjacent partitions. The mathematical form of the equation is:

[0088]

[0089] In the formula:

[0090] is the optical diffusion coefficient, and its value is directly related to the light guide plate material, LED layout and light source characteristics;

[0091] is the Laplace operator of the brightness distribution, simulating the diffusion effect of brightness in the spatial dimension;

[0092] is the liquid crystal molecule deflection voltage response function (Sigmoid function); its derivative reflects the non - linear influence of voltage change on the deflection rate of liquid crystal molecules;

[0093] is the gain coefficient, used to adjust the driving intensity of the input brightness on the backlight brightness. This equation quantifies the brightness diffusion coupling relationship between adjacent partitions.

[0094] Subsequently, establish a multi - objective optimization problem to mathematize the trade - off relationship between contrast, power consumption and response time. The objective function is defined as:

[0095]

[0096] In the formula:

[0097] , representing the mean square error between the target brightness and the actual brightness;

[0098] is the power consumption of the driving circuit, which is proportional to the square of the backlight brightness and the driving voltage;

[0099] is the liquid crystal response time, which is determined by the change rate of the driving voltage and the material characteristics;

[0100] , , is the dynamic weight coefficient, which is adjusted in real time through the data of the temperature sensor.

[0101] Further define the constraint conditions to ensure that the optimization results meet the requirements of physical realizability and visual smoothness:

[0102] 1. Minimum brightness constraint:

[0103]

[0104] Where is the maximum drivable brightness of the partition , which is determined by the hardware specifications of the backlight module.

[0105] 2. Adjacent partition brightness difference constraint:

[0106]

[0107] In the formula represents the set of adjacent partitions of the partition , and this constraint acts synergistically with the term in the brightness diffusion equation to suppress the halo effect.

[0108] Finally, an improved particle swarm optimization algorithm is used to solve the optimization problem, and the steps are as follows:

[0109] Step 1: Particle swarm initialization

[0110] Define the particle position vector , representing the brightness allocation scheme of the -th particle for the partitions. Exemplarily, the particle swarm size is set to 50, and the initial position of each particle is randomly generated within the allowable range of partition brightness:

[0111]

[0112] In the formula is a random number uniformly distributed in [0, 1], which is determined by the lowest brightness constraint.

[0113] The generation range of the initial particles and the dynamic partition grid in step S2 ensure that the brightness values in each partition comply with the physical limitations of their minimum controllable area size.

[0114] Step 2: Fitness calculation

[0115] Calculate the fitness value of each particle according to the aforementioned objective function:

[0116]

[0117] The weight coefficient is adjusted in real time according to the formula where the temperature parameter is obtained by the sensor in step S1.

[0118] Apply a penalty term to the particles that violate the brightness difference constraint:

[0119]

[0120] In the formula is the penalty coefficient (exemplary value is 1000), and this constraint condition is associated with the partition merging rule in step S2.

[0121] Step 3: Update of individual and global optimal

[0122] Record the historical optimal position of each particle and the global optimal position of the population , and the update rule is:

[0123]

[0124]

[0125] Step 4: Update of particle velocity and position

[0126] Adjust the velocity and position according to the improved particle swarm update formula:

[0127]

[0128]

[0129] The inertia weight is adjusted dynamically according to the formula where is the real-time temperature collected in step S1. When the temperature rises, reduce to suppress particle oscillation and improve convergence stability.

[0130] Velocity term The allowable range is related to the aforementioned diffusion coefficient and expands the search range in the region with strong diffusion ability ( large value):

[0131]

[0132] In the formula is the diffusion coefficient of the partition which is calculated from the refractive index of the light guide plate and the LED pitch

[0133] Step 5: Constraint forced processing

[0134] Perform the following forced correction on the updated particle positions:

[0135] 1. Boundary constraint:

[0136] If then set ;

[0137] If then set

[0138] 2. Brightness difference constraint:

[0139] For each partition traverse its adjacent partitions If then scale the brightness value proportionally:

[0140]

[0141] The set of adjacent partitions is derived from the dynamic partition grid generated in step S2 to ensure that the constraint conditions are consistent with the spatial relationship of the physical partitions.

[0142] Step 6: Termination condition judgment

[0143] Repeat steps 2 - 5 until any of the following termination conditions are met:

[0144] The change range of the population optimal fitness value is less than 1% in 20 consecutive iterations;

[0145] The total number of iterations reaches the preset upper limit (exemplarily set to 200 times).

[0146] When the temperature reduce the convergence judgment threshold (e.g., adjust from 1% to 2%) to avoid the algorithm falling into local optimum in a high - temperature environment.​​

[0147] Step 7: Output the optimized result

[0148] Use the globally optimal position as the optimal brightness value for each partition Output it and transmit it to the gamma mapping module.

[0149] In this embodiment, through the deep coupling of the physical model and the optimization algorithm, the global optimization of the backlight brightness distribution is realized.

[0150] For step S4, its specific implementation process is as follows:

[0151] First, based on the optimal brightness value of each partition output in step S3 , generate a non-linear gamma mapping curve for each partition. The curve is obtained by fitting the error function that minimizes the error between the target brightness and the actual brightness. The error function is defined as:

[0152]

[0153] In the formula, is the driving voltage sampling point (exemplary values are from 0 to 5V, with a step of 0.1V), is the gamma curve function of partition , is the voltage corresponding measured brightness value.

[0154] Furthermore, a piecewise quadratic function is used to locally fit the gamma curve to balance the computational complexity and the fitting accuracy. The form of the piecewise function is:

[0155]

[0156] The piecewise interval is dynamically divided according to the optimal brightness value . The specific rule is:

[0157] Divide the voltage range corresponding to equally into sub-intervals (exemplary value is ), and a quadratic function is independently fitted for each sub-interval.

[0158] Subsequently, the input brightness is converted to the driving voltage through inverse function calculation. For each pixel point , solve according to the gamma curve of the partition to which it belongs:

[0159]

[0160] In the formula, Represents the inverse function of the gamma curve, solved using the Newton iteration method. The iteration formula is:

[0161]

[0162] Initial iteration value Set as the driving voltage value of the corresponding partition in the previous frame to improve the calculation efficiency.

[0163] For step S5, its specific implementation process is as follows:

[0164] First, the matrix operations of the luminance diffusion model established in step S3 and the iterative calculations of the optimization algorithm are executed in parallel by a hardware accelerator. The hardware accelerator can be implemented using FPGA or ASIC. Exemplarily, the hardware accelerator is implemented using FPGA, and its architecture includes a parallel computing unit and a pipeline scheduling module.

[0165] The parallel computing unit processes the luminance diffusion equation of the backlight partition in blocks, specifically realizing the discretization of the finite difference method of the luminance diffusion model:

[0166]

[0167] In the formula, is the spatial step size, is the time step size, is obtained in real time from the liquid crystal response derivative data pre-stored in the look-up table.

[0168] Furthermore, the pipeline scheduling module dynamically allocates computing resources according to the real-time requirements of the partitions. Exemplarily, the highest priority is given to high refresh rate regions (such as the center region of the game screen), and its scheduling strategy satisfies:

[0169]

[0170] Where is the temporal change rate of the target luminance of the partition, is the diffusion coefficient defined in the luminance diffusion model.

[0171] Subsequently, according to the actual luminance data fed back by the optical sensor , the diffusion coefficient and the optimization weight parameters of the luminance diffusion model are dynamically adjusted. The adjustment formula is:

[0172]

[0173]

[0174] In the formula is the total number of partitions, is the maximum luminance of the backlight module, and They respectively represent the L1 and L2 norms. The adjustment is executed in real time through the interrupt service program of the hardware accelerator.

[0175] After the driving signal is synthesized, it is output to the backlight driving circuit through the LVDS interface. At the same time, the adjusted parameters are written back to the optimization model in step S3 to form a closed-loop control link.

[0176] Generally speaking, the present invention first dynamically generates a non-uniform backlight partition grid based on the brightness characteristics of the input image and environmental parameters. This partitioning rule ensures independent control of the detail areas through the entropy value threshold and brightness difference constraints. Subsequently, a brightness diffusion model is established in combination with the physical characteristics of the backlight module, and a temperature-adaptive multi-objective optimization algorithm is used to solve the optimal brightness of each partition. This process realizes real-time solution through the dynamic binding of the particle swarm algorithm and hardware parameters. Then, a non-linear gamma mapping curve is generated for each partition, and the brightness signal is accurately mapped to the driving voltage through inverse function conversion. Finally, the hardware accelerator executes the computing tasks in parallel, and dynamically adjusts the diffusion coefficient and optimization weight based on the feedback data of the optical sensor to form a complete control link from parameter acquisition, model optimization to driving output. Cross-layer collaboration is realized between each step through the real-time data stream and parameter write-back mechanism to ensure the physical adaptability of brightness control and the real-time response ability of dynamic scenes.

[0177] The liquid crystal display dynamic contrast optimization system described below can be correspondingly referred to the liquid crystal display dynamic contrast optimization method described above.

[0178] Please refer to the attached Figure 2 , the present invention also provides a liquid crystal display dynamic contrast optimization system, including:

[0179] A data acquisition module 10, used to obtain the image brightness component and environmental parameters;

[0180] A dynamic partitioning module 20, generating a non-uniform backlight partition grid based on the image entropy value;

[0181] A collaborative optimization module 30, establishing a brightness diffusion model and solving a multi-objective optimization problem;

[0182] A gamma mapping module 40, generating a non-linear gamma curve and synthesizing a driving signal;

[0183] An execution feedback module 50, executing the drive through the hardware accelerator and adjusting the parameters in a closed loop.

[0184] The system of this embodiment can be used to execute the above method embodiment, and its principle and technical effect are similar, so it will not be elaborated here.

[0185] Although embodiments of the present invention have been shown and described, it will be understood by those of ordinary skill in the art that various changes, modifications, substitutions and variations can be made to these embodiments without departing from the principles and spirit of the present invention, and the scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. A method for optimizing dynamic contrast of a liquid crystal display screen, characterized in that: The following steps are involved: Collect the brightness component and environmental parameters of the input image; Dynamically generate a non-uniform backlight partition grid based on image content; Establish a backlight brightness collaborative optimization model to solve the optimal brightness value of each partition; generating a nonlinear gamma mapping curve according to the optimal brightness value; Synthesize drive signals and perform closed-loop feedback adjustments; The step of dynamically generating a non-uniform backlight partition grid based on image content comprises: The local entropy value of each area of ​​the image is calculated by sliding window traversal, and the image is divided into multiple initial partitions according to the preset entropy value threshold, where the high entropy value area corresponds to the detail-rich area; Dynamically adjust the entropy value threshold according to the ambient light intensity, merge adjacent partitions whose areas are smaller than the preset threshold, and generate a final non-uniform backlight partition grid; The step of establishing a backlight brightness collaborative optimization model and solving the optimal brightness value of each partition includes: A brightness diffusion model is built based on the refractive index of the light guide plate of the backlight module and the LED spacing parameters to simulate the brightness diffusion relationship between adjacent partitions; Constructing a multi-objective optimization model, wherein the multi-objective optimization model takes maximizing contrast, minimizing power consumption and minimizing liquid crystal response time as optimization objectives, and introduces a brightness difference constraint condition between adjacent partitions; The step of generating a nonlinear gamma mapping curve according to the optimal brightness value comprises: According to the optimal brightness value of each partition, a nonlinear gamma mapping curve is generated for each partition, wherein the gamma mapping curve is obtained by fitting a function that minimizes an error between target brightness and actual brightness; Based on the proportional relationship between the input brightness and the target brightness, the input brightness is mapped into a driving voltage signal; The steps of synthesizing the driving signal and performing closed-loop feedback adjustment include: The matrix operations of the brightness diffusion model and the iterative calculations of the optimization algorithm are performed in parallel through the hardware accelerator; According to the actual brightness data fed back by the light sensor, the diffusion coefficient of the brightness diffusion model is dynamically adjusted and the weight parameters are optimized.

2. The method for optimizing dynamic contrast ratio of a liquid crystal display according to claim 1, characterized in that: The specific rules for merging adjacent partitions whose areas are smaller than a preset threshold include: Determine the difference in the mean brightness of adjacent partitions, and merge them if the difference is less than the preset tolerance range; The area of ​​the merged partitions does not exceed the minimum controllable area size of the backlight module.

3. The method for optimizing dynamic contrast ratio of a liquid crystal display screen according to claim 1, characterized in that: The method for solving the multi-objective optimization model includes: An improved particle swarm algorithm is used for iterative calculation, and the inertia weight of the algorithm is dynamically adjusted according to the real-time temperature of the backlight module; During the particle update process, the brightness difference between adjacent partitions is constrained not to exceed a preset threshold.

4. The method for optimizing dynamic contrast ratio of a liquid crystal display screen according to claim 1, characterized in that: The method for generating the nonlinear gamma mapping curve comprises: The gamma curve of each partition is locally fitted using a piecewise quadratic function, where the segment interval is dynamically divided according to the optimal brightness value of the partition; The input brightness value is converted into a driving voltage through an inverse function calculation.

5. The method for optimizing dynamic contrast ratio of a liquid crystal display screen according to claim 1, characterized in that: The hardware accelerator is implemented using FPGA or ASIC, and its architecture includes: A parallel computing unit for processing the brightness diffusion equation of the backlight partition in blocks; The pipeline scheduling module allocates computing resources to partitions with high real-time requirements according to priority.

6. A liquid crystal display screen dynamic contrast optimization system, used to execute the method according to any one of claims 1 to 5, characterized in that: include: A data acquisition module is used to obtain image brightness components and environmental parameters; Dynamic partitioning module, which generates non-uniform backlight partitioning grid based on image entropy value; Collaborative optimization module, which builds brightness diffusion model and solves multi-objective optimization problems; Gamma mapping module, which generates nonlinear gamma curve and synthesizes driving signal; The execution feedback module executes the drive and adjusts the parameters in a closed loop through the hardware accelerator.

Images