Flexible Transform Kernel Selection for Video Block Compression
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional video codecs limit transform kernel selection based on block size, leading to suboptimal compression performance and higher bit rates due to the inability to choose the most efficient transform for a given block of video data.
Innovation Solution
Implement a flexible transform selection mechanism that allows independent selection of transform kernel types for the long and short sides of a transform block, accommodating a wider range of transform types for various block sizes.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If transform kernel selection is limited based on block size in conventional codecs, then device complexity is reduced and ease of operation is improved, but compression efficiency deteriorates and bit rate increases
Solution Approach 1:
The transform block is divided into two independent dimensions: long side and short side. Each side can independently select its transform kernel type (DCT or Identity), allowing flexible combination of transform types across different block sizes without requiring a completely new selection mechanism.
Solution Approach 2:
The transform kernel selection becomes dynamic and adaptive based on the actual block dimensions. Instead of static rules that limit kernel choices, the system dynamically determines appropriate transform types for each side of the block, enabling optimal compression performance for varying block sizes.
2Productivity
If transform kernel selection is limited based on block size, then ease of operation is improved, but compression efficiency deteriorates
Solution Approach 1:
By segmenting the transform selection into independent long-side and short-side choices, the system achieves both simplicity and efficiency. Each side can be selected independently based on its characteristics, allowing optimal compression without complex multi-dimensional selection rules.
Solution Approach 2:
Different transform kernel types can be applied to different sides of the same transform block based on local characteristics. This allows the encoder to apply DCT transforms where they provide better compression while using Identity transforms where they are more appropriate, optimizing compression efficiency for each region.
3Loss of information
If transform kernel selection is restricted by block size conventions, then device complexity is reduced, but bit rate increases
Solution Approach 1:
The transform selection mechanism dynamically adapts to block size characteristics without requiring complex predefined rules. By independently selecting transform types for long and short sides based on actual dimensions, the system minimizes information loss and maintains video quality while avoiding overly complex selection logic.
Data Source
AI summary
A long transform type to apply to a long side of a transform block and a short transform type to apply to a short side of the transform block are identified. Identifying the long transform type and the short transform type includes determining that the long side is equal to a first threshold value; and, in response to determining that the long side is equal to the first threshold value, coding the long transform type, the long transform type being one of a discrete cosine transform or an identity transform. The long transform type and the short transform type are then applied.