Video Encoder Transform Basis Selection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional video encoding methods require significant processing to find the best combination of transform basis and transform parameters, leading to high computational load without optimizing compression efficiency.
Innovation Solution
The proposed encoder and decoder systems determine whether to perform first and second transforms based on intra prediction modes and transform bases, allowing for the skipping of unnecessary transforms and reducing processing load while maintaining compression efficiency.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Manufacturing precision
If conventional video encoding methods search for the best combination of transform basis and transform parameters, then compression efficiency is improved, but computational load increases significantly
Solution Approach 1:
The patent applies parameter changes by dynamically selecting transform bases and transform types based on intra prediction modes. Specifically, when the intra prediction mode is planar or DC mode, the patent selects specific transform bases (e.g., DST-VII or DCT-II) and determines whether to perform second transforms based on block size and prediction mode combinations. This conditional parameter selection avoids exhaustive searching while maintaining compression efficiency.
Solution Approach 2:
The patent performs preliminary action by determining the transform basis and transform type before actually performing the transforms. The encoder first identifies the intra prediction mode, then pre-determines which transform basis to use and whether second transforms are necessary, based on predetermined rules. This preliminary determination eliminates the need for exhaustive searching during the encoding process, reducing computational load while preserving compression efficiency.
2Manufacturing precision
If exhaustive search for optimal transform parameters is performed, then compression efficiency is maximized, but processing time increases
Solution Approach 1:
The patent changes parameters by establishing predetermined relationships between intra prediction modes and transform bases. Instead of exhaustively searching all possible transform parameter combinations, the system uses the intra prediction mode (planar, DC, or directional) to directly determine the appropriate transform basis and whether second transforms should be applied. This parameter mapping approach significantly reduces processing time while maintaining compression efficiency.
Solution Approach 2:
The patent performs preliminary action by pre-establishing rules that map intra prediction modes to specific transform configurations. Before encoding, the system determines which transform basis to use and whether second transforms are needed based on the predicted intra mode. This preliminary determination eliminates time-consuming exhaustive searches during the actual encoding process.
3Manufacturing precision
If multiple transform bases and parameters are evaluated, then compression efficiency improves, but device complexity increases
Solution Approach 1:
The patent simplifies device complexity by changing the approach from exhaustive evaluation of multiple transform bases to a deterministic selection process. The intra prediction mode directly determines which transform basis to use and whether second transforms are applicable. This reduces encoding complexity while maintaining compression efficiency through intelligent parameter selection based on prediction mode characteristics.
Solution Approach 2:
The patent performs preliminary action by establishing predetermined rules that eliminate the need for complex evaluation of multiple transform bases. The system pre-determines the transform configuration based on intra prediction mode, block size, and position, avoiding the need for complex real-time evaluation of multiple transform options during encoding.
Data Source
AI summary
An encoder which encodes a current block of a picture includes a processor and memory. Using the memory, the processor: determines whether intra prediction is to be used for the current block; and when it is determined that intra prediction is to be used for the current block, generates first transform coefficients by performing first transform of residual signals of the current block using a first transform basis; quantizes the first transform coefficients when an intra prediction mode for the current block is a determined mode and the first transform basis is different from a determined transform basis; and generates second transform coefficients by performing second transform of the first transform coefficients using a second transform basis, and quantizes the second transform coefficients, when the intra prediction mode for the current block is not the determined mode or when the first transform basis matches the determined transform basis.