Chroma Intra Prediction Block Correction Using Color Correlation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional image encoding and decoding methods lack the improvement in handling the correlation between color components, leading to performance issues and block degradation during image processing.
Innovation Solution
An image encoding/decoding method that addresses the correlation between color components, by generating prediction blocks based on color component correlation and applying corrections to these blocks to improve prediction accuracy and reduce block degradation.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional image encoding/decoding methods are used, then the processing is simpler, but prediction accuracy is low and block degradation occurs
Solution Approach 1:
The current block is divided into multiple prediction blocks, and different correction settings are applied to different regions. The correction process is segmented into multiple steps: generating initial prediction blocks, determining correction settings based on block size and prediction mode, applying corrections to specific regions, and combining results. This segmentation allows for improved prediction accuracy while managing complexity through structured processing.
Solution Approach 2:
The patent applies preliminary correction actions to prediction blocks before final reconstruction. Correction settings are determined in advance based on block size and prediction mode characteristics. Reference pixels are pre-processed and corrections are applied to prediction blocks before they are combined with residual blocks, ensuring that accuracy improvements are built into the encoding/decoding process early rather than as a post-processing step.
2Reliability
If correction is applied to prediction blocks, then block degradation is reduced, but processing complexity increases
Solution Approach 1:
Different correction settings are applied to different regions and types of blocks based on their specific characteristics. The patent determines correction settings according to block size (e.g., 4×4, 8×8, 16×16) and prediction mode type (intra prediction, inter prediction). This local quality approach ensures that each block receives appropriate correction treatment, improving overall reliability while avoiding unnecessary complexity for blocks that don't require extensive correction.
Solution Approach 2:
The patent modifies prediction blocks by applying parameter-based corrections derived from color component correlations. Correction parameters are calculated based on the relationship between different color components (Y, Cb, Cr) and applied to adjust prediction block values. This parameter change approach systematically improves block quality through mathematical transformations rather than complex procedural operations.
3Productivity
If color component correlation is utilized, then encoding performance is improved, but computational requirements increase
Solution Approach 1:
The patent uses color component correlation information that is already available in the encoding/decoding system for multiple purposes: generating prediction blocks, determining correction settings, and improving reconstruction quality. By making the color component correlation analysis multi-functional, the patent extracts maximum encoding performance benefit from a single computational source, reducing the need for separate computational processes.
Solution Approach 2:
The encoding/decoding system uses its own internal color component data to generate correction information, rather than requiring external computational resources. The color components (Y, Cb, Cr) that are already being processed for color representation are also utilized to generate prediction and correction data, allowing the system to serve its own encoding needs without additional computational overhead.
Data Source
AI summary
Disclosed is an image decoding method using the correlation between color components to perform into prediction of chrominance components. Here, the image decoding method using the correlation between color components to perform intra prediction of chrominance components comprises the steps of: checking image data and a prediction mode in a bitstream; generating a prediction block according to a reconstructed prediction mode; determining compensation settings according to the size of a current block and the reconstructed prediction mode; compensating the prediction block according to the determined compensation settings; and reconstructing the current block by adding reconstructed image data to the prediction block.