Image Processing Method for LCD Color Cast and Grid Feeling
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing image processing technologies for liquid crystal displays struggle to optimize both color cast and grid feeling simultaneously, leading to suboptimal image display quality.
Innovation Solution
An image processing method that acquires Gaussian probabilities for preset colors in a scene, identifies relevant picture regions, and applies color output correction values using a Gaussian model and brightness adjustment, effectively addressing color cast and grid feeling issues through viewing angle compensation.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If viewing angle compensation technology is applied to correct color cast, then color cast is improved, but grid feeling increases and image display quality deteriorates
Solution Approach 1:
The patent applies different processing strategies to different picture regions. It identifies regions containing preset colors (such as skin tones, blue skies, grass) and applies Gaussian probability-based correction only to these specific regions, while leaving other regions unchanged. This localized approach corrects color cast in key areas without causing global grid feeling across the entire image.
Solution Approach 2:
The patent uses a Gaussian probability model to dynamically determine correction intensity for different colors. By calculating the probability that a detected color belongs to a preset color category, it applies variable correction strength - stronger correction for high-probability matches and weaker or no correction for low-probability matches. This parameter-based control enables selective color cast correction without excessive grid feeling.
2Object-generated harmful factors
If one driving mode is used for viewing angle compensation, then picture quality is improved, but color cast correction effect is worse
Solution Approach 1:
The patent creates a dynamic correction system that adapts to different driving modes and image content. Rather than using a fixed correction table, it calculates Gaussian probabilities in real-time based on the actual color distribution in the image and the selected driving mode. This allows the system to optimize correction strength dynamically - using stronger correction when needed and weaker correction when picture quality should be preserved.
Solution Approach 2:
The system incorporates feedback through the Gaussian probability calculation, which continuously monitors the color distribution in the image and adjusts correction parameters accordingly. By comparing detected colors against preset color models and calculating probability scores, the system receives feedback about which colors need correction and by how much, enabling adaptive optimization of both picture quality and color cast correction.
3Measurement precision
If color correction is applied to all colors, then color cast is reduced, but similar colors in different scenes may be misdetected
Solution Approach 1:
The patent segments the color correction process by identifying and isolating specific picture regions that contain preset colors. It uses color space analysis to distinguish between colors that belong to preset categories (such as skin tones in portraits, blue skies, grass) and colors that do not. This segmentation allows targeted correction of relevant colors while ignoring or preserving similar colors in inappropriate contexts, preventing misdetection.
Solution Approach 2:
The patent replaces simple threshold-based color detection with a probabilistic Gaussian model. Instead of using fixed thresholds that can be easily fooled by similar colors, it uses statistical probability calculations to determine whether a detected color belongs to a preset category. This substitution of mechanical thresholding with statistical reasoning enables more accurate color identification and prevents misdetection of similar colors in different scenes.
Data Source
AI summary
The present application discloses an image processing method, an image processing device, and a computer device. The image processing method includes: acquiring a first to-be-processed chromaticity data set and a second to-be-processed chromaticity data set of a preset color of a to-be-processed picture, acquiring a Gaussian probability of the preset color, identifying a picture region in the to-be-processed picture that contains the preset color of the preset scene, and acquiring a color output correction value. The method of the present application can effectively improve a color cast phenomenon and a quality of the displayed picture.