Intra Prediction Reference Blocks for Texture-Aware Video Coding
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing video coding and decoding standards suffer from poor video coding and decoding performance due to inadequate intra prediction methods, which fail to effectively utilize spatial correlation for improved compression efficiency.
Innovation Solution
The proposed methods involve determining a first intra prediction value for a current block based on sample values of neighboring reconstructed areas and a reference block, and optionally using multiple reference blocks to enhance the accuracy of intra prediction, thereby improving coding efficiency and reducing bitstream overhead.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If traditional intra prediction methods are used, then the coding process is simple, but the video coding and decoding performance is poor
Solution Approach 1:
The current block is divided into multiple sub-blocks, and different prediction methods are applied to different sub-blocks. This segmentation allows the system to achieve better overall prediction performance by combining multiple simpler prediction approaches rather than using a single complex method across the entire block.
Solution Approach 2:
The patent extends the prediction approach by utilizing multiple reference blocks from different spatial locations and directions, not just traditional single reference blocks. This multi-dimensional reference approach enhances prediction accuracy by capturing spatial correlations from various orientations.
2Measurement precision
If multiple reference blocks are used to improve prediction accuracy, then the intra prediction accuracy improves, but the computational complexity and bitstream overhead increase
Solution Approach 1:
Different prediction methods are applied to different regions within the current block based on local characteristics. By identifying and applying appropriate prediction strategies to specific sub-blocks with similar local properties, the system achieves high prediction accuracy without uniformly applying complex methods across the entire block, thus reducing overall computational complexity.
Solution Approach 2:
The patent applies multiple reference blocks and prediction methods selectively to sub-blocks where they provide the most benefit, rather than uniformly applying them to the entire current block. This partial application strategy achieves sufficient prediction accuracy while minimizing the increase in computational complexity and bitstream overhead.
3Productivity
If more reference blocks are utilized, then the spatial correlation is better exploited, but the bitstream overhead increases
Solution Approach 1:
Multiple reference blocks are used selectively for different sub-blocks rather than for the entire current block. This partial application allows the system to exploit spatial correlation where beneficial while avoiding the bitstream overhead of transmitting reference information for all blocks, thus improving compression efficiency without excessive overhead.
Solution Approach 2:
The current block is segmented into sub-blocks, and reference block information is transmitted only for sub-blocks where multiple references provide value. This segmentation strategy improves compression efficiency by exploiting spatial correlation in regions where it exists while minimizing bitstream overhead by avoiding unnecessary reference information transmission in other regions.
Data Source
AI summary
A decoding method, a coding method, a decoder and a coder are provided. The decoding method includes: determining a first reference block of the current block; obtaining a first intra-frame prediction value of the current block according to a sample value of a reconstructed area adjacent to the current block and a sample value of the first reference block; and determining a reconstruction value of the current block according to the first intra-frame prediction value of the current block.