Subblock Transform Decoding for Video Coding Efficiency
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing video coding technologies face challenges in efficiently encoding and decoding video data, particularly in effectively utilizing subblock transform (SBT) modes to enhance transform coding efficiency.
Innovation Solution
The proposed method involves a system and method for decoding a sub-block transform (SBT) coding unit, where the SBT coding unit is partitioned into multiple transform units. The method includes receiving syntax elements from a bitstream to identify non-zero transform coefficients, determining the transform units with non-zero coefficients, and decoding these coefficients while setting other transform units to zero.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If subblock transform (SBT) modes are used to enhance transform coding efficiency, then compression performance is improved, but computational complexity increases
Solution Approach 1:
The transform unit is divided into multiple subblocks, and different transform types (DST-VII and DCT-II) are selectively applied to different subblocks based on the presence of non-zero coefficients. This segmentation allows the system to process only relevant subblocks, improving compression performance while managing computational complexity by avoiding unnecessary transforms on zero-coefficient subblocks.
Solution Approach 2:
Different transform types are applied to different subblocks within the same transform unit based on local characteristics (presence of non-zero coefficients). This local quality approach optimizes compression for each subblock independently, applying DST-VII to subblocks with non-zero coefficients and DCT-II to others, thereby improving overall compression performance while maintaining manageable computational complexity.
2Reliability
If all transform units are decoded regardless of zero coefficients, then decoding completeness is ensured, but unnecessary computations increase
Solution Approach 1:
The patent extracts and processes only the relevant subblocks that contain non-zero transform coefficients, while setting subblocks with zero coefficients to zero without performing full decoding operations. This extraction approach ensures decoding completeness for non-zero regions while eliminating unnecessary computations for zero regions, thereby improving decoding efficiency without compromising reliability.
Solution Approach 2:
Instead of performing complete decoding operations on all transform units, the patent applies partial action by selectively decoding only those subblocks that contain non-zero coefficients. This partial action approach maintains decoding completeness for relevant data while significantly reducing unnecessary computations on zero-coefficient subblocks, thus improving decoding efficiency.
Data Source
AI summary
A computing device performs a method of decoding video data by receiving, from a bitstream, a first syntax element associated with an SBT coding unit that indicates that there exists at least one non-zero transform coefficient in the SBT coding unit; determining a first transform unit that includes non-zero transform coefficients; receiving a second syntax element associated with a first chroma component of the first transform unit, a third syntax element associated with a second chroma component, and a fourth syntax element associated with luma component of the first transform unit; decoding the transform coefficients of chroma and luma components of the first transform unit, based on the second, third and fourth syntax elements; and setting transform coefficients of luma and chroma components of rest of the plurality of transform units to zeros.