Video Encoding and Decoding via Adaptive Residual Sub-Block Transforms
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing video encoding and decoding technologies using sub-block transform (SBT) are not flexible enough to adapt to various residual distribution situations, affecting encoding and decoding efficiency.
Innovation Solution
The coding block is further partitioned to obtain a target sub-block, with entropy decoding and inverse quantization performed on transform sub-blocks to generate a reconstructed residual, inferring residuals in other areas as zero, adapting to diverse residual distributions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If a coding block is partitioned into sub-blocks using traditional SBT technology, then transform processing can be performed, but the method is not flexible enough to adapt to various residual distribution situations
Solution Approach 1:
The patent divides the coding block into multiple sub-blocks (first sub-block, second sub-block, third sub-block, fourth sub-block) and applies different transform processing strategies to each sub-block based on residual distribution characteristics. This segmentation allows the system to adapt to various residual patterns while maintaining manageable complexity through systematic processing of each segment.
Solution Approach 2:
The patent applies different transform processing methods to different sub-blocks based on their local residual characteristics. Specifically, it performs transform processing on sub-blocks with significant residual energy while skipping transform on sub-blocks with minimal residuals, thereby adapting locally to residual distribution patterns and improving overall efficiency.
2Reliability
If transform processing is performed on all sub-blocks, then transform performance is maintained, but encoding and decoding efficiency decreases
Solution Approach 1:
The patent performs transform processing only on necessary sub-blocks (those with significant residual energy) while skipping transform on sub-blocks where it is unnecessary. This partial application of transform processing maintains adequate reconstruction quality while significantly improving encoding and decoding efficiency by avoiding redundant operations.
Solution Approach 2:
The patent dynamically adjusts the transform processing parameter (whether to perform transform) based on residual distribution characteristics of each sub-block. By changing this binary parameter adaptively, the system maintains transform performance where needed while improving overall encoding efficiency through selective application.
Data Source
AI summary
In a method for video decoding, a target sub-block on which entropy decoding is to be performed is obtained, from a bitstream, based on partitioning a coding block according to a residual block corresponding to the coding block. Transform sub-block information about at least one transform sub-block is obtained, the at least one transform sub-block obtained by partitioning the target sub-block. An inverse quantization coefficient sub-block corresponding to the at least one transform sub-block is obtained based on entropy decoding and inverse quantizing the at least one transform sub-block with the transform sub-block information about the at least one transform sub-block. A reconstructed residual corresponding to the coding block is generated based on inverse transforming the inverse quantization coefficient sub-block corresponding to the at least one transform sub-block, a residual of an area other than the target sub-block in the coding block being inferred to be zero.