Closed-loop simulation evaluation method and system for backlight diffusion model
The closed-loop simulation evaluation method and system based on the backlight diffusion model solves the problem that existing technologies cannot evaluate the visual error of approximate algorithms in high-precision physical systems. It provides a quantitative evaluation tool, reduces computational complexity while maintaining high visual fidelity, and quantifies the relationship between computational load and image quality loss.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-12-16
- Publication Date
- 2026-03-31
AI Technical Summary
Existing technologies lack an evaluation scheme that can simulate the "closed-loop" effect, making it impossible to assess the final visual error of approximation algorithms when applied to high-precision physical systems. Furthermore, they ignore the differences between image processing and physical optics, and cannot quantify the relationship between reduced computational complexity and image quality loss.
A closed-loop simulation evaluation method and system for backlight diffusion model is proposed. By converting the image domain through gamma decoding and encoding, and combining the calculation of partitioned backlight value and liquid crystal value, fitted and reference images are generated. The color difference formula is used to quantify the perceived difference and evaluate the image quality.
It achieves closed-loop evaluation of fast approximation models in real physical systems, provides physically correct quantitative evaluation that conforms to human visual perception, reduces computational complexity while maintaining high visual fidelity, and quantifies the relationship between computational cost and image quality loss.
Smart Images
Figure CN121767253A_ABST
Abstract
Description
Technical Field
[0001] This application relates to the field of display system technology, and in particular to a closed-loop simulation evaluation method and system for a backlight diffusion model. Background Technology
[0002] In modern high-performance display systems (such as local dimming systems using Mini-LED backlighting), achieving high dynamic range (HDR) and high image quality requires precise calculation of light propagation at the physical level and accurate compensation of the liquid crystal (LC). To achieve this high-precision compensation, the industry typically employs high-fidelity physical optical models. These models (often called "reference models") typically combine measured light spread function (LSF / PSF) and pixel-level flat field correction (FFC) data. Their implementation usually involves large-scale 2D convolution and access to high-resolution FFC lookup tables. While these high-precision models provide a reliable "gold standard" reference, they place enormous computational and storage burdens on display controllers (such as TCONs). For example, 2D convolution is computationally extremely expensive, and it also requires storing a large LSF core and a full-resolution FFC map. Therefore, the industry urgently needs to develop "fast approximation models" with lower computational and storage overhead to achieve acceptable image quality in cost-sensitive TCON hardware.
[0003] In developing the aforementioned "fast approximation model," a key technical problem arises: how to scientifically and quantitatively evaluate the "visual fidelity" of these approximation models? Since in real display systems, the intermediate results (e.g., liquid crystal compensation values) generated by the fast model (e.g., for backlight modeling) ultimately affect the actual physical display hardware, simply comparing the output of the "fast model" with the output of the "baseline model" is insufficient. Summary of the Invention
[0004] The main objective of this application is to propose a closed-loop simulation evaluation method and system for backlight diffusion models, so as to fully and effectively evaluate the backlight diffusion models.
[0005] To achieve the above objectives, one aspect of this application proposes a closed-loop simulation evaluation method for a backlight diffusion model, the method comprising the following steps: The input image in sRGB format is converted into a linear light domain image by gamma decoding; wherein the linear light domain image is used as the initial linear image. The partition backlight value is calculated based on the initial linear image; The fitted screen backlight is calculated based on the partitioned backlight values; the fitted liquid crystal values are calculated based on the fitted screen backlight and the initial linear image. A reference screen backlight is generated based on the partitioned backlight values; a reference liquid crystal value is calculated based on the reference screen backlight and the initial linear image; The fitted liquid crystal value is multiplied by the backlight of the reference screen to generate a fitted linear image; The reference liquid crystal value and the reference screen backlight are multiplied together to generate a reference linear image; The fitted linear image and the reference linear image are converted into sRGB format by gamma encoding to obtain the fitted sRGB image and the reference sRGB image respectively. By comparing the fitted sRGB image and the reference sRGB image, the target perceptual difference is obtained; Evaluate the image quality of the fitted sRGB image.
[0006] In some embodiments, generating a reference screen backlight based on the partitioned backlight value includes the following steps: The low-resolution partition backlight value Mapping to high-resolution sparse mesh The calculation formula is: ; Where H represents the number of vertical backlight zones, W represents the number of horizontal backlight zones; and the δ function is the impulse function. Will Compared with the measured 2D physics LSF core Perform 2D convolution to obtain The calculation formula is: ; in, This represents the luminous effect of the backlight after it has been diffused. Use full-resolution FFC mapping Perform calibration The backlight of the reference screen is obtained. The calculation formula is: .
[0007] In some embodiments, multiplying the reference liquid crystal value and the reference screen backlight to generate a reference linear image includes the following steps: The reference liquid crystal value compensated by LC and the reference screen front backlight Multiply to obtain the reference linear image. ; Wherein, the reference liquid crystal value for LC compensation is: .
[0008] In some embodiments, calculating the fitted front backlight based on the partitioned backlight values includes the following steps: The low-resolution partition backlight value Expanded into a blocky high-resolution image The calculation formula is: ; Where Sr represents the number of pixels covered in the horizontal direction of each backlight partition, and Sc represents the number of pixels covered in the vertical direction of each backlight partition. High-resolution images With an optimized 2D separable Gaussian kernel Convolution is performed to obtain the fitted screen backlight; wherein, the 2D separable Gaussian kernel Fitting has physical Mapping with FFC The combination of .
[0009] In some embodiments, calculating the fitted liquid crystal value based on the fitted screen backlight and the initial linear image includes the following steps: The fitted liquid crystal value is calculated based on the fitted screen backlight and the initial linear image, using the following formula: ; in, Represents the transmittance of the liquid crystal; Represents the brightness of the backlight. The value represents the brightness of the input image, (x,y) represents the pixel coordinates, and the superscript wavy line represents the estimated value.
[0010] In some embodiments, comparing the fitted sRGB image and the reference sRGB image to obtain the target perceptual difference includes the following steps: The target perceptual difference is obtained by comparing the fitted sRGB image and the reference sRGB image using a color difference formula used to quantify the perceptual difference between two colors.
[0011] In some embodiments, evaluating the image quality of the fitted sRGB image includes the following steps: The image quality of the fitted sRGB image is evaluated based on the peak signal-to-noise ratio.
[0012] To achieve the above objectives, another aspect of this application proposes a closed-loop simulation evaluation system for a backlight diffusion model, the system comprising: An image decoding unit is used to convert an input image in sRGB format into a linear light domain image through gamma decoding; wherein the linear light domain image serves as an initial linear image. A backlight partitioning unit is used to calculate the partitioned backlight value based on the initial linear image; The fitting value calculation unit is used to calculate the fitting screen front backlight based on the partition backlight value; and to calculate the fitting liquid crystal value based on the fitting screen front backlight and the initial linear image. A reference value calculation unit is used to generate a reference screen backlight based on the partitioned backlight value; and to calculate a reference liquid crystal value based on the reference screen backlight and the initial linear image. The fitted image generation unit is used to multiply the fitted liquid crystal value and the backlight of the reference screen to generate a fitted linear image. A reference image generation unit is used to multiply the reference liquid crystal value and the backlight in front of the reference screen to generate a reference linear image; An image encoding unit is used to convert the fitted linear image and the reference linear image into sRGB format respectively through gamma encoding, so as to obtain a fitted sRGB image and a reference sRGB image. A difference sensing unit is used to compare the fitted sRGB image and the reference sRGB image to obtain the target perceived difference; An image evaluation unit is used to evaluate the image quality of the fitted sRGB image.
[0013] To achieve the above objectives, another aspect of this application provides an electronic device, which includes a memory and a processor. The memory stores a computer program, and the processor executes the computer program to implement the above-described method.
[0014] To achieve the above objectives, another aspect of the embodiments of this application proposes a computer-readable storage medium storing a computer program that, when executed by a processor, implements the above-described method.
[0015] To achieve the above objectives, another aspect of this application provides a computer program product, including a computer program that, when executed by a processor, implements the above-described method.
[0016] The embodiments of this application include at least the following beneficial effects: This application provides a closed-loop simulation evaluation method and system for a backlight diffusion model. The scheme converts an input image in sRGB format into a linear optical domain image using gamma decoding; the linear optical domain image serves as the initial linear image; partitioned backlight values are calculated based on the initial linear image; a fitted front-screen backlight is calculated based on the partitioned backlight values; fitted liquid crystal values are calculated based on the fitted front-screen backlight and the initial linear image; a reference front-screen backlight is generated based on the partitioned backlight values; a reference liquid crystal value is calculated based on the reference front-screen backlight and the initial linear image; the fitted liquid crystal value and the reference front-screen backlight are multiplied to generate a fitted linear image; the reference liquid crystal value and the reference front-screen backlight are multiplied to generate a reference linear image; the fitted linear image and the reference linear image are converted to sRGB format using gamma encoding to obtain a fitted sRGB image and a reference sRGB image; the fitted sRGB image and the reference sRGB image are compared to obtain the target perception difference; and the image quality of the fitted sRGB image is evaluated. This application achieves closed-loop evaluation by calculating the perceptual difference between the fitted sRGB image and the reference sRGB image and evaluating the image quality of the fitted sRGB image, which can fully evaluate the performance of the fast approximation model that generates the fitted sRGB image. Attached Figure Description
[0017] To more clearly illustrate the technical solutions in the embodiments of this application, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.
[0018] Figure 1 A schematic flowchart of a closed-loop simulation evaluation method for a backlight diffusion model provided in an embodiment of this application; Figure 2 An example flowchart of a closed-loop simulation evaluation method for a backlight diffusion model provided in this application embodiment; Figure 3 A comparison example of a reference image and a fitted image provided for an embodiment of this application; Figure 4 Another example of comparison between a reference image and a fitted image provided for an embodiment of this application; Figure 5 A schematic diagram of the structure of a closed-loop simulation evaluation system for a backlight diffusion model provided in this application embodiment; Figure 6 This is a schematic diagram of the hardware structure of an electronic device provided in an embodiment of this application. Detailed Implementation
[0019] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative of this application and are not intended to limit it. In the following description, when referring to the accompanying drawings, unless otherwise indicated, the same numbers in different drawings represent the same or similar elements. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with those of this application; they are merely examples of systems and methods consistent with some aspects of the embodiments of this application as detailed in the appended claims.
[0020] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application belongs. The terminology used herein is for the purpose of describing embodiments of this application only and is not intended to limit this application.
[0021] Before providing a detailed description of the embodiments of this application, some related technologies involved in the embodiments of this application will be described first, as follows: Terminology Explanation: TCON (Timing Controller): The core chip in the display system, responsible for receiving image data and controlling the control panel driver.
[0022] LSF (Light Spread Function): The light spread function. It describes the physical properties of light rays emanating from a point source and spreading across a display panel.
[0023] PSF (Point Spread Function): Similar to LSF, it describes the diffusion characteristics of a point light source.
[0024] FFC (Flat-Field Correction): A technique used to compensate for inherent brightness or color uniformity in display panels (such as the Mura effect).
[0025] LC (Liquid Crystal): A display medium used to control the amount of light passing through it.
[0026] (DeltaE2000): A color difference formula used to quantify the perceived difference between two colors. The smaller the value, the more difficult it is for the human eye to distinguish the difference.
[0027] sRGB: A standard color space widely used in monitors, printers, and the Internet.
[0028] MAC (Multiply-Accumulate Operation): Multiply-accumulate operation. The basic unit for measuring computational complexity.
[0029] Golden Reference Model: Refers to a high-precision, high-complexity model whose simulation results are considered "ground truth" or "gold standard" and are used to measure the accuracy of other models.
[0030] FittedModeling / Algorithm-Under-Test: Refers to a faster approximation algorithm with lower computational cost, designed to replace the baseline model.
[0031] FSC (Field Sequential Color): A display technology that synthesizes a color image by rapidly and sequentially displaying the three color fields (R / G / B) without using color filters.
[0032] PSNR (Peak Signal-to-Noise Ratio): A standard for measuring image quality.
[0033] Existing technologies lack evaluation schemes capable of simulating the "closed-loop" effect. Specifically, existing technologies suffer from the following technical problems: 1. Lack of closed-loop verification mechanism: It is impossible to evaluate the final visual error generated when the compensation parameters (such as LC compensation value) calculated by an approximate algorithm (such as a fast backlight model) are actually applied to a high-precision physical system (real display optical characteristics).
[0034] 2. Improper physical domain processing: Many evaluation methods ignore the differences between image processing (e.g., sRGB gamma space) and physical optics (linear optical domain). Physical light superposition must be calculated in the linear domain, while perceptual evaluation (e.g., sRGB) must be performed in the perceptual domain. Incorrect domain processing will lead to distorted evaluation results.
[0035] 3. Inability to quantify trade-offs: Developers need a way to clearly quantify the relationship between "reduced computational complexity" (e.g., fewer MACs) and "image quality loss" (e.g., increased) in order to make the best engineering decisions for TCON hardware design.
[0036] This application aims to solve the above problems and provide a general, end-to-end display algorithm evaluation system and method. This application can simulate the closed-loop performance of the algorithm in a real physical system with high fidelity and provide a physically correct quantitative evaluation that conforms to human visual perception.
[0037] This application provides a closed-loop simulation evaluation method and system for a backlight diffusion model, relating to the field of display system technology. The closed-loop simulation evaluation method and system for a backlight diffusion model provided in this application can be applied to a terminal, a server, or software running on a terminal or server. In some embodiments, the terminal can be a smartphone, tablet, laptop, desktop computer, smart speaker, smartwatch, or in-vehicle terminal, but is not limited to these; the server can be configured as an independent physical server, a server cluster composed of multiple physical servers, or a distributed system; it can also be configured as a cloud server providing basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communication, middleware services, domain name services, security services, CDN, and big data and artificial intelligence platforms; the server can also be a node server in a blockchain network; the software can be an application implementing a closed-loop simulation evaluation method for a backlight diffusion model, but is not limited to the above forms.
[0038] This application can be used in a wide variety of general-purpose or special-purpose computer system environments or configurations. Examples include: personal computers, server computers, handheld or portable devices, tablet devices, multiprocessor systems, microprocessor-based systems, set-top boxes, programmable consumer electronics devices, network PCs, minicomputers, mainframe computers, and distributed computing environments including any of the above systems or devices. This application can be described in the general context of computer-executable instructions executed by a computer, such as program modules. Generally, program modules include routines, programs, objects, components, data structures, etc., that perform specific tasks or implement specific abstract data types. This application can also be practiced in distributed computing environments where tasks are performed by remote processing devices connected via a communication network. In distributed computing environments, program modules can reside in local and remote computer storage media, including storage devices.
[0039] Reference Figure 1 This application provides a closed-loop simulation evaluation method for a backlight diffusion model. This method may include, but is not limited to, steps S100 to S180, as follows: S100: Convert the input image in sRGB format into a linear light domain image through gamma decoding; wherein, the linear light domain image serves as the initial linear image; S110: Calculate the partition backlight value based on the initial linear image; S120: Calculate the fitted front backlight based on the partitioned backlight value; calculate the fitted liquid crystal value based on the fitted front backlight and the initial linear image; S130: Generate a reference screen backlight based on the partitioned backlight value; calculate a reference liquid crystal value based on the reference screen backlight and the initial linear image; S140: Multiply the fitted liquid crystal value and the backlight of the reference screen to generate a fitted linear image; S150: Multiply the reference liquid crystal value and the reference screen backlight to generate a reference linear image; S160: Convert the fitted linear image and the reference linear image into sRGB format respectively by gamma encoding to obtain the fitted sRGB image and the reference sRGB image; S170: Compare the fitted sRGB image and the reference sRGB image to obtain the target perception difference; S180: Evaluate the image quality of the fitted sRGB image.
[0040] Optionally, generating a reference screen backlight based on the partitioned backlight values includes the following steps: The low-resolution partition backlight value Mapping to high-resolution sparse mesh The calculation formula is: ; Where H represents the number of vertical backlight zones, W represents the number of horizontal backlight zones; and the δ function is the impulse function. Will Compared with the measured 2D physics LSF core Perform 2D convolution to obtain The calculation formula is: ; in, This represents the luminous effect of the backlight after it has been diffused. Use full-resolution FFC mapping Perform calibration The backlight of the reference screen is obtained. The calculation formula is: .
[0041] Optionally, multiplying the reference liquid crystal value and the backlight of the reference screen to generate a reference linear image includes the following steps: The reference liquid crystal value compensated by LC and the reference screen front backlight Multiply to obtain the reference linear image. ; Wherein, the reference liquid crystal value for LC compensation is: .
[0042] Optionally, calculating the fitted front backlight based on the partitioned backlight values includes the following steps: The low-resolution partition backlight value Expanded into a blocky high-resolution image The calculation formula is: ; Where Sr represents the number of pixels covered in the horizontal direction of each backlight partition, and Sc represents the number of pixels covered in the vertical direction of each backlight partition. High-resolution images With an optimized 2D separable Gaussian kernel Convolution is performed to obtain the fitted screen backlight; wherein, the 2D separable Gaussian kernel Fitting has physical Mapping with FFC The combination of .
[0043] Optionally, calculating the fitted liquid crystal value based on the fitted screen backlight and the initial linear image includes the following steps: The fitted liquid crystal value is calculated based on the fitted screen backlight and the initial linear image, using the following formula: ; in, Represents the transmittance of the liquid crystal; Represents the brightness of the backlight. The value represents the brightness of the input image, (x,y) represents the pixel coordinates, and the superscript wavy line represents the estimated value.
[0044] Optionally, comparing the fitted sRGB image and the reference sRGB image to obtain the target perceptual difference includes the following steps: The target perceptual difference is obtained by comparing the fitted sRGB image and the reference sRGB image using a color difference formula used to quantify the perceptual difference between two colors.
[0045] Optionally, evaluating the image quality of the fitted sRGB image includes the following steps: The image quality of the fitted sRGB image is evaluated based on the peak signal-to-noise ratio.
[0046] The following sections will provide a detailed description and explanation of some optional embodiments of this application, using specific application examples.
[0047] This embodiment proposes an end-to-end evaluation system and method for evaluating approximation algorithms in display systems. The system is specifically designed to quantify the perceptual error introduced by a "fast model to be evaluated" relative to a "high-precision benchmark physical model." The core idea of the system is to establish two parallel processing paths and simulate the closed-loop effect of the algorithm through a specific "cross-driven" logic, finally comparing the errors in a unified perceptual domain.
[0048] Figure 2 This document presents an example flowchart of a closed-loop simulation evaluation method for a backlight diffusion model according to this embodiment. The process comprises two parallel paths: 1. Upper part: FittedModelingPath. This represents a fast approximation model implemented in TCON. It receives a "linearized image" and a "partitioned backlight," and outputs "fitted liquid crystal values" (i.e., intermediate parameters). 2. Lower part: GoldenReferencePath. This represents a high-precision physical simulation that simulates the "true values" of the display hardware. It also receives a "linearized image" and a "local dimming value," and generates a "reference front-facing backlight" (reference physical response) through "physical optical simulation" (such as LSF convolution and FFC). Core cross-logic: The key to evaluating the system is that the "fitted liquid crystal values" in the upper part are multiplied by the "reference front-facing backlight" in the lower part to obtain a "fitted linear image." Final comparison: Both the "fitted linear image" and the "reference linear image" are converted to the sRGB domain, and then the perceived error (e.g., ...) is calculated through the "objective quality assessment" module. ).
[0049] Specifically, the solution in this embodiment includes the following modules and steps: 1. Data Input and Preprocessing Module: Source Image (sRGB): Provides a standard input image in the sRGB color space. Linear Image: The input image must first be converted into a physically linear light domain image through gamma decoding. Localized Backlight: Calculates the local dimming value (i.e., the localized backlight value) based on the "linear image". This "localized backlight" value will serve as the common input for subsequent parallel paths.
[0050] 2. Parallel processing paths: The "zoned backlight" value is simultaneously fed into two paths. The "linear image" is also used by two paths to calculate the liquid crystal value.
[0051] Path A: The algorithm path to be evaluated, which implements the computationally efficient "fast approximation model" that the user wants to evaluate. Step A-1: Calculate the "fitted front-screen backlight" based solely on the "zone backlight" value. Step A-2: Calculate the "fitted liquid crystal value" based on the "fitted front-screen backlight" and "linear image" from Step A-1.
[0052] Path B: Reference Path. This path implements a high-fidelity "physical simulation model" representing the "true ground value." Step B-1: Based solely on the "zone backlight" value, generate the "reference screen backlight," i.e., the reference physical response, through high-precision calculation. Step B-2: Based on the "reference screen backlight" and "linear image" from Step B-1, calculate the "reference liquid crystal value."
[0053] 3. Closed-Loop Simulation and Output Generation Module (Core of this Embodiment): This module is crucial to this embodiment. The evaluation logic is not a simple comparison of the backlight profiles of the two paths. Step C-1 (Post-Modeling Output / Cross-Drive): Multiply the "fitted liquid crystal value" (Step A-2) generated by path A with the "reference front-screen backlight" (Step B-1) generated by path B. The result generates a "fitted linear image." This output realistically simulates the linear light emitted by the display if "Fast Model A" is deployed in "Real Physical System B." Step C-2 (Reference Output): Multiply the "reference liquid crystal value" (Step B-2) generated by path B with the "reference front-screen backlight" (Step B-1) generated by path B. The result generates a "reference linear image." This represents the image that should ideally be displayed.
[0054] 4. Evaluation and Comparison Module: Step D-1 (Gamma Encoding): The "fitted linear image" generated in Step C-1 and the "reference linear image" generated in Step C-2 are gamma encoded and converted back to sRGB space to obtain the "fitted sRGB image" and the "reference sRGB image". Step D-2 (Comparison): Finally, the "objective index comparison" module is used to compare the "fitted sRGB image" and the "reference sRGB image". This step can use various objective image quality evaluation indicators, such as PSNR (Peak Signal-to-Noise Ratio). When it is necessary to evaluate perceptual differences, it is preferable to use a color difference formula that conforms to human visual perception, such as... This is used to quantitatively calculate the perceived differences between the two. Through this evaluation framework, selected indicators (such as...) The value of ) precisely quantifies the final perceived error in the real physical system introduced by the "fast approximation model A".
[0055] A more specific implementation method is as follows: The evaluation system in this embodiment can be used to evaluate a "fast backlight modeling method based on separable Gaussian filtering" and compare it with a "traditional high-precision model based on 2D convolution". Specifically, it includes the following technical solutions: 1. Path B (Benchmark Reference Path): Implemented as a "traditional high-precision model".
[0056] Step B-1: "Reference Screen Front Backlight" for this path Generate through the following steps: 1. Point source modeling: This involves analyzing low-resolution zoned backlight values. Mapping to high-resolution sparse mesh The calculation formula is: ; Specifically, H represents the number of vertical backlight zones; W represents the number of horizontal backlight zones.
[0057] For example, if the backlight partition is 9x16, then (H,W)=(9,16).
[0058] The delta function is an impulse function, which has a value only at the origin. if (x = 0, y = 0) :δ(x,y) = 1; else δ(x,y) = 0.
[0059] The purpose of the above calculation is to construct a sparse grid and place the LEDs. For example, suppose we need to adjust the LED value I. BL If (0,0) is placed at x=25, y=30 in a sparse grid, then x0=25, y0=30; therefore, the above calculation formula at coordinates (25,30) becomes: I src (25,30) = I BL (0,0)*δ(25-x0, 30-y0) =I BL (0,0)*δ(0,0) =I BL (0,0).
[0060] 2. Optical convolution: [This refers to...] Compared with the measured 2D physics LSF core Perform 2D convolution on a pixel (e.g., 1301x1301 pixels) to obtain... The calculation formula is: ; in, The Isrc represents the backlight's luminous effect after diffusion; the Isrc in front represents a display panel with distributed backlight, such as a display panel with a 1920*1080 resolution. If the backlight is divided into 16*9 zones, then there will be an LED emitting light in the center of each zone. Placing 16*9 LEDs in a 1920*1080 grid will result in a sparse Isrc.
[0061] Convolution of the LED panel with LSF represents the LED emitting light. The light from the emitting LED is distributed to nearby pixels according to the shape of Kphy, so each pixel will have some "brightness". The matrix that records the brightness of each pixel is Iconv.
[0062] 3. Flattening Correction: Use full-resolution FFC mapping (For example, 3840x2160 pixels) are corrected to obtain the final "reference screen front backlight". The calculation formula is: ; Step B-2: "Reference Linear Output" through ideal LC compensation and Multiplying together yields (i.e.) ).
[0063] II. Path A (Algorithm path to be evaluated): Implemented as "Proposed Fast Gaussian Approximation Model".
[0064] Step A-1: Approximate Backlight Profile of the Path Generate through the following steps: 1. Block upsampling: Using zero-order hold, the low-resolution local backlight values are upsampled. Expanded into a blocky high-resolution image The calculation formula is: ; Where Sr represents the number of pixels covered horizontally in each backlight zone, and Sc represents the number of pixels covered vertically in each backlight zone. For example, if a backlight covers a 25x25 display area, then Sr=Sc=25.
[0065] Under the above settings, I BL Each value will fill 25x25 pixels, meaning the pixel values in that backlit area are equal to the backlight value. Iup is a matrix of resolution size, such as 1920x1080; while I... BL It is a matrix representing the number of backlight zones, such as 16*9.
[0066] 2. Separable Gaussian Filter: ... With an optimized 2D separable Gaussian kernel Perform convolution. Use the Gaussian kernel. It approximates physics. Mapping with FFC The combined effect.
[0067] Step A-2: The "intermediate parameter" (i.e., the compensated LC) is calculated as follows : ; Among them, the above formula is consistent with the aforementioned formula. Correspondingly, both involve calculating liquid crystal values. Represents the transmittance of the liquid crystal; Represents the brightness of the backlight. The value represents the brightness of the input image, (x, y) represents the pixel coordinates, and the superscript tilde represents the estimated value. For the pixel at (x, y), the display shows: Final light emission = Backlight brightness * Liquid crystal transmittance; Therefore, by changing the formula, the liquid crystal value can be calculated as follows: Liquid crystal transmittance = final light emission (input image) / backlight brightness.
[0068] The above formula is the same as the one mentioned above. The difference is I act A wavy line has been added above the actual value. The superscript wavy line represents the estimated value, while the value without a superscript wavy line is the true value, which is the estimated value. Further estimates Therefore, a wavy line was added above the liquid crystal value.
[0069] Cross-drive and evaluation (steps C and D): Step C: Develop path A... The path generated by path B Multiply by the above to obtain the "modeled linear output". Step D: Multiply the above output by the "baseline linear output" (i.e., Convert each to sRGB color space and calculate. Color difference.
[0070] Environment Selection for this Implementation: The evaluation system in this implementation is preferably based on the Field Sequential Color (FSC) model. The key advantage of this model is that the FSC system does not use a traditional color filter; its R, G, and B color fields are displayed independently and sequentially. This mechanism provides a purer optical environment, and its light diffusion (LSF) characteristics are not affected by factors such as sub-pixel crosstalk. Therefore, this scenario can more accurately and without interference evaluate the simulation fidelity of the backlight model's physical effects on light diffusion, thereby reducing the interference of other variables on the evaluation results. Experimental Configuration and Results of this Implementation: Tests were conducted using the UHD-IQA database on a 4K (3840x2160) resolution, 48x42 backlight partition FSC Mini-LED model. Path B (Traditional Model) uses... The LSF core requires approximately [amount] per frame (RGB). MACs operation. Path A (Gaussian model) uses... Equivalent 1D core Each frame only requires MACs calculation. Evaluation results: This evaluation system (step D-3) measured that the computational load of path A is reduced by approximately 1800 times compared to path B, while maintaining high visual fidelity, with an average color difference of only [missing information]. This embodiment demonstrates that the evaluation system of this embodiment successfully quantifies the trade-off between computation and accuracy in fast algorithms.
[0071] Reference Figure 3 The comparison between the reference image (left sub-image) and the fitted image (middle sub-image) with the fast fitting path parameter sigma=10 in the experimental example demonstrates that the experimental method can reflect the fitting results in a way that conforms to human visual perception. The right sub-image shows the difference between the real image and the synthesized image, highlighting the effect of the algorithm used. Under this setting, the following results were obtained: The effect.
[0072] Reference Figure 4 The comparison between the reference image (left sub-image) and the fitted image (middle sub-image) when the fast fitting path parameter sigma=120 in the experimental example demonstrates that the experimental method can reflect the fitting results in a way that conforms to human visual perception. Under this setting, the following results were obtained. In this case, the indicator is better and the display effect is better. We can reverse the evaluation of the effect of the parameter sigma and select the optimal parameter.
[0073] The beneficial effects of this embodiment include: 1. Closed-loop verification is achieved, resulting in more realistic evaluation results: This embodiment solves the problem that existing technologies cannot evaluate the closed-loop effect. By using compensation parameters generated by the "algorithm to be evaluated" (such as...) ) "Drives" high-precision "physical reference models" (such as This system can accurately simulate the final visual effect produced by the approximation algorithm on real physical hardware, and the evaluation results are closer to the actual application scenario.
[0074] 2. Ensures physical correctness and high evaluation accuracy: This embodiment emphasizes the correctness of the processing domain. The system mandates that all physical optical simulations (such as convolution and superposition) be performed in the "linear optical domain" and the final quality assessment (such as...) be performed in the "perceptual domain" (sRGB). This rigorous domain management ensures the scientific rigor and accuracy of the evaluation results.
[0075] 3. A universal evaluation framework is provided, with wide applicability: The system architecture of this embodiment is universal. The "algorithm path to be evaluated" and the "benchmark reference path" are pluggable modules. Developers can use this framework to evaluate and compare any algorithm aimed at approximating backlight diffusion effects, making it highly applicable.
[0076] 4. Quantified performance trade-offs to guide hardware design: This system provides a quantitative tool that enables TCON chip designers to weigh performance trade-offs between computational complexity (e.g., MACs) and perceived image quality loss (e.g., ... A clear, data-driven trade-off must be made between these factors. For example, the implementation demonstrates that reducing computation by 1800 times only results in an average color difference of 1.655, providing ample justification for adopting this approximation algorithm.
[0077] 5. Controllable evaluation environment, reduced interference: By selecting specific display models such as FSC (in the embodiments), an evaluation environment with purer physical characteristics (e.g., no subpixel crosstalk) can be constructed. This allows the method to more accurately separate and evaluate the modeling fidelity of specific physical effects (e.g., light diffusion) while eliminating interference from other factors.
[0078] Reference Figure 5 This application also provides a closed-loop simulation evaluation system for a backlight diffusion model, which can implement the above-mentioned closed-loop simulation evaluation method for a backlight diffusion model. The system includes: An image decoding unit is used to convert an input image in sRGB format into a linear light domain image through gamma decoding; wherein the linear light domain image serves as an initial linear image. A backlight partitioning unit is used to calculate the partitioned backlight value based on the initial linear image; The fitting value calculation unit is used to calculate the fitting screen front backlight based on the partition backlight value; and to calculate the fitting liquid crystal value based on the fitting screen front backlight and the initial linear image. A reference value calculation unit is used to generate a reference screen backlight based on the partitioned backlight value; and to calculate a reference liquid crystal value based on the reference screen backlight and the initial linear image. The fitted image generation unit is used to multiply the fitted liquid crystal value and the backlight of the reference screen to generate a fitted linear image. A reference image generation unit is used to multiply the reference liquid crystal value and the backlight in front of the reference screen to generate a reference linear image; An image encoding unit is used to convert the fitted linear image and the reference linear image into sRGB format respectively through gamma encoding, so as to obtain a fitted sRGB image and a reference sRGB image. A difference sensing unit is used to compare the fitted sRGB image and the reference sRGB image to obtain the target perceived difference; An image evaluation unit is used to evaluate the image quality of the fitted sRGB image.
[0079] It is understood that the content of the above method embodiments is applicable to this system embodiment. The specific functions implemented in this system embodiment are the same as those in the above method embodiments, and the beneficial effects achieved are also the same as those achieved in the above method embodiments.
[0080] This application also provides an electronic device, which includes a memory and a processor. The memory stores a computer program, and the processor executes the computer program to implement the method of this application. This electronic device can be any smart terminal, including tablet computers, in-vehicle computers, etc.
[0081] It is understood that the content of the above method embodiments is applicable to the device embodiments. The specific functions implemented by the device embodiments are the same as those of the methods of this application, and the beneficial effects achieved are the same as those achieved by the methods of this application.
[0082] Figure 6 The hardware structure of an electronic device according to another embodiment is illustrated. The electronic device includes: The processor 101 can be implemented using a general-purpose CPU (Central Processing Unit), microprocessor, application-specific integrated circuit (ASIC), or one or more integrated circuits, and is used to execute relevant programs to implement the technical solutions provided in the embodiments of this application. The memory 102 can be implemented as a read-only memory (ROM), static storage device, dynamic storage device, or random access memory (RAM). The memory 102 can store the operating system and other applications. When the technical solutions provided in the embodiments of this specification are implemented through software or firmware, the relevant program code is stored in the memory 102 and is called and executed by the processor 101. Input / output interface 103 is used to implement information input and output; The communication interface 104 is used to enable communication and interaction between this device and other devices. Communication can be achieved through wired means (such as USB, network cable, etc.) or wireless means (such as mobile network, WIFI, Bluetooth, etc.). Bus 105 transmits information between various components of the device (e.g., processor 101, memory 102, input / output interface 103, and communication interface 104); The processor 101, memory 102, input / output interface 103 and communication interface 104 are connected to each other within the device via bus 105.
[0083] This application also provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the method of this application.
[0084] It is understood that the content of the above method embodiments is applicable to this storage medium embodiment. The specific functions implemented in this storage medium embodiment are the same as those in the above method embodiments, and the beneficial effects achieved are also the same as those achieved in the above method embodiments.
[0085] This application also provides a computer program product, including a computer program that, when executed by a processor, implements the above-described method.
[0086] Memory, as a non-transitory computer-readable storage medium, can be used to store non-transitory software programs and non-transitory computer-executable programs. Furthermore, memory may include high-speed random access memory, and may also include non-transitory memory, such as at least one disk storage device, flash memory device, or other non-transitory solid-state storage device. In some embodiments, memory may optionally include memory remotely located relative to the processor, and these remote memories can be connected to the processor via a network. Examples of such networks include, but are not limited to, the Internet, intranets, local area networks, mobile communication networks, and combinations thereof.
[0087] The embodiments described in this application are for the purpose of more clearly illustrating the technical solutions of the embodiments of this application, and do not constitute a limitation on the technical solutions provided by the embodiments of this application. As those skilled in the art will know, with the evolution of technology and the emergence of new application scenarios, the technical solutions provided by the embodiments of this application are also applicable to similar technical problems.
[0088] Those skilled in the art will understand that the technical solutions shown in the figures do not constitute a limitation on the embodiments of this application, and may include more or fewer steps than shown, or combine certain steps, or different steps.
[0089] The system embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs.
[0090] Those skilled in the art will understand that all or some of the steps in the methods disclosed above, as well as the functional modules / units in the systems and devices, can be implemented as software, firmware, hardware, or suitable combinations thereof.
[0091] The terms “first,” “second,” “third,” “fourth,” etc. (if present) in the specification and accompanying drawings of this application are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of this application described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms “comprising” and “having,” and any variations thereof, are intended to cover non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.
[0092] It should be understood that in this application, "at least one (item)" means one or more, and "more than" means two or more. "And / or" is used to describe the relationship between related objects, indicating that three relationships can exist. For example, "A and / or B" can represent three cases: only A exists, only B exists, and both A and B exist simultaneously, where A and B can be singular or plural. The character " / " generally indicates that the preceding and following related objects are in an "or" relationship. "At least one (item) of the following" or similar expressions refer to any combination of these items, including any combination of single or plural items. For example, at least one (item) of a, b, or c can represent: a, b, c, "a and b", "a and c", "b and c", or "a and b and c", where a, b, and c can be single or multiple.
[0093] In the embodiments provided in this application, it should be understood that the disclosed systems and methods can be implemented in other ways. For example, the system embodiments described above are merely illustrative; for instance, the division of the units described above is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection between systems or units may be electrical, mechanical, or other forms.
[0094] The units described above as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.
[0095] Furthermore, the functional units in the various embodiments of this application can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit.
[0096] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes multiple instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods of the various embodiments of this application. The aforementioned storage medium includes various media capable of storing programs, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.
[0097] The preferred embodiments of the present application have been described above with reference to the accompanying drawings, but this does not limit the scope of the claims of the present application. Any modifications, equivalent substitutions, and improvements made by those skilled in the art without departing from the scope and substance of the embodiments of the present application shall be within the scope of the claims of the present application.
Claims
1. A method for closed loop simulation evaluation of a backlight diffusion model, characterized in that, The method comprises the following steps: Converting an input image in sRGB format into a linear light domain image by gamma decoding; wherein the linear light domain image is an initial linear image; Calculating a partitioned backlight value based on the initial linear image; Calculating a fitted front-of-screen backlight based on the partitioned backlight value; calculating a fitted liquid crystal value based on the fitted front-of-screen backlight and the initial linear image; Generating a reference front-of-screen backlight based on the partitioned backlight value; calculating a reference liquid crystal value based on the reference front-of-screen backlight and the initial linear image; Multiplying the fitted liquid crystal value and the reference front-of-screen backlight to generate a fitted linear image; Multiplying the reference liquid crystal value and the reference front-of-screen backlight to generate a reference linear image; Converting the fitted linear image and the reference linear image into sRGB format by gamma encoding respectively to obtain a fitted sRGB image and a reference sRGB image; Comparing the fitted sRGB image and the reference sRGB image to obtain a target perceptual difference; Evaluating the image quality of the fitted sRGB image.
2. The method of claim 1, wherein, The method for generating a reference front-of-screen backlight based on the partitioned backlight value comprises the following steps: mapping the low resolution partitioned backlight values to a high resolution sparse grid , the calculation being: ； Wherein, H represents the number of vertical backlight partitions, W represents the number of horizontal backlight partitions; the delta function is an impulse function; will be described below. with the measured 2D physical LSF core performing a 2D convolution, resulting in with the measured 2D physical LSF core performing a 2D convolution, resulting in with the measured 2D physical LSF core ； wherein, represents the light emitting effect after the backlight is diffused; Using full resolution ffc mapping Correcting obtaining the reference on-screen backlight The calculation is 。 3. The method of claim 2, wherein, The method for multiplying the reference liquid crystal value and the reference front-of-screen backlight to generate a reference linear image comprises the following steps: The reference liquid crystal value compensated by LC is multiplied by the reference front-of-screen backlight to obtain the reference linear image ; Wherein, the reference liquid crystal value compensated by LC is: 。 4. The method of claim 1, wherein, The method for calculating a fitted front-of-screen backlight based on the partitioned backlight value comprises the following steps: low resolution partition backlight values extended to a block-shaped high resolution image , the calculation formula is: ； Wherein, Sr represents the number of pixels covered by each backlight partition in the horizontal direction, Sc represents the number of pixels covered by each backlight partition in the vertical direction; high resolution image with an optimized 2D separable Gaussian kernel is convolved to obtain the fitted frontlight; wherein the 2D separable Gaussian kernel is fitted with a physical and FFC mapping combination.
5. The method of claim 4, wherein, The method for calculating a fitted liquid crystal value based on the fitted front-of-screen backlight and the initial linear image comprises the following steps: The calculation formula for calculating a fitted liquid crystal value based on the fitted front-of-screen backlight and the initial linear image is: ； wherein represents the transmittance of the liquid crystal; represents the luminance of the backlight, represents the luminance of the input image, (x, y) represents the pixel coordinates, and the superscript tilde represents an estimated value.
6. The method of claim 1, wherein, The method for comparing the fitted sRGB image and the reference sRGB image to obtain a target perceptual difference comprises the following steps: Comparing the fitted sRGB image and the reference sRGB image by using a color difference formula for quantifying the perceptual difference between two colors to obtain the target perceptual difference.
7. The method of claim 1 to 6, wherein, The method for evaluating the image quality of the fitted sRGB image comprises the following steps: Evaluating the image quality of the fitted sRGB image according to the peak signal-to-noise ratio.
8. A closed loop simulation evaluation system for a backlight diffusion model, characterized by, The system comprises: An image decoding unit configured to convert an input image in sRGB format into a linear light domain image by gamma decoding; wherein the linear light domain image is an initial linear image; A backlight partitioning unit configured to calculate a partitioned backlight value based on the initial linear image; A fitting value calculation unit configured to calculate a fitted front-of-screen backlight based on the partitioned backlight value; and calculate a fitted liquid crystal value based on the fitted front-of-screen backlight and the initial linear image; A reference value calculation unit configured to generate a reference front-of-screen backlight based on the partitioned backlight value; and calculate a reference liquid crystal value based on the reference front-of-screen backlight and the initial linear image; a fitting image generating unit configured to multiply the fitting liquid crystal value and the reference pre-screen backlight to generate a fitting linear image; a reference image generating unit configured to multiply the reference liquid crystal value and the reference pre-screen backlight to generate a reference linear image; an image encoding unit configured to convert the fitting linear image and the reference linear image into sRGB format respectively by gamma encoding to obtain a fitting sRGB image and a reference sRGB image; a difference perception unit configured to compare the fitting sRGB image and the reference sRGB image to obtain a target perceived difference; an image evaluation unit configured to evaluate image quality of the fitting sRGB image.
9. An electronic device, comprising: The electronic device comprises a memory and a processor, the memory stores a computer program, and the processor implements the method in any one of claims 1 to 7 when executing the computer program.
10. A computer-readable storage medium storing a computer program, the computer-readable storage medium comprising: The computer program is executed by the processor to implement the method in any one of claims 1 to 7.