Local Illumination Compensation Linear Model for Video Encoding
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current video encoding and decoding methods face challenges in efficiently handling local illumination changes, leading to increased complexity and potential visual artifacts due to the complexity of computing Local Illumination Compensation (LIC) parameters.
Innovation Solution
The method involves simplifying the computation of LIC parameters using a linear model based on sums of absolute differences of neighboring reconstructed and reference samples, with co-location determined by an L-shape adjacent to the predicted block, and incorporating a regularization process to prevent discontinuity issues when using multiple models.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If complex LIC parameter computation methods are used, then illumination compensation accuracy is improved, but computational complexity increases
Solution Approach 1:
The patent changes the computational parameters from complex multi-step calculations to a simplified linear model using only slope and intercept parameters. This is achieved by using co-located samples from an L-shaped region and computing parameters through direct linear regression rather than complex iterative optimization, thereby reducing computational complexity while maintaining compensation accuracy
Solution Approach 2:
The patent extracts only the essential components needed for LIC (slope and intercept parameters from co-located samples) and discards complex computational steps. By focusing solely on the linear relationship between reference and current block samples, the method removes unnecessary computational overhead while preserving the core illumination compensation function
2Measurement precision
If multiple LIC models are used, then compensation accuracy is improved, but visual artifacts increase due to discontinuity
Solution Approach 1:
The patent applies different linear models to different spatial regions (current block vs. reference block) using co-located samples from the L-shaped region. Each region gets its own locally optimized slope and intercept parameters, ensuring accurate local illumination compensation while maintaining continuity across block boundaries through the standardized parameter derivation process
3Manufacturing precision
If standard LIC computation is used, then encoding accuracy is improved, but encoding speed decreases
Solution Approach 1:
The patent performs preliminary computation of LIC parameters using co-located samples from the L-shaped region before actual encoding. By pre-calculating the slope and intercept parameters from readily available reference and current block samples, the method prepares compensation data in advance, enabling faster encoding without sacrificing accuracy
Solution Approach 2:
The patent replaces complex iterative mechanical computation systems with a direct linear algebraic solution. Instead of using complex optimization algorithms that require multiple iterations, the method substitutes a closed-form linear regression calculation that computes parameters in a single step, dramatically improving encoding speed while maintaining precision
Data Source
AI summary
In at least one embodiment, a method and apparatus for encoding/decoding a picture comprising predicting at least one block, wherein the predicting comprises performing motion compensation and local illumination compensation based on a reference block, the local illumination compensation including applying a linear model based on sums of absolute differences of neighboring reconstructed samples and corresponding reference samples of the reference block, wherein the neighboring reconstructed samples and corresponding reference samples of the reference block are co-located according to an L-shape substantially adjacent to the block to be predicted, the L-shape comprising a row of pixels located to the top side of the predicted block and a column of pixels located to the left side of the predicted block, the co-location being determined according a motion vector of the predicted block.