Secondary Transform Matrix Selection for Video Encoding
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current video processing techniques face challenges in efficiently managing bandwidth demand for digital video, particularly as the number of connected devices increases, leading to growing bandwidth requirements.
Innovation Solution
The proposed method involves context modeling for residual coding in video processing, specifically by applying a secondary transform tool to video blocks based on characteristics such as intra prediction direction, coding tools, and color components, thereby optimizing encoding and decoding processes.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of energy
If a secondary transform tool is applied to all video blocks, then compression performance is improved, but computational complexity and processing time increase
Solution Approach 1:
The patent applies the secondary transform tool selectively to specific video blocks based on local characteristics such as prediction mode, block size, and gradient information. Rather than uniformly applying the transform to all blocks, the system adapts the transform application to local regions where it provides the most benefit, thereby improving compression efficiency while limiting the increase in processing complexity to only necessary areas.
Solution Approach 2:
The patent introduces multiple parameters to control secondary transform application including block size thresholds, gradient magnitude thresholds, and prediction mode conditions. By dynamically adjusting these parameters based on content characteristics, the system optimizes the balance between compression performance and processing complexity, applying the computationally intensive secondary transform only when parameters indicate it will be beneficial.
2Manufacturing precision
If multiple transform matrices are selected and applied, then encoding precision is improved, but encoding time increases
Solution Approach 1:
The patent segments the set of available transform matrices into different categories or groups based on their characteristics and suitability for different video content types. Rather than evaluating all possible matrices for every block, the system divides the search space into segments and selects from relevant segments based on block characteristics, thereby maintaining encoding precision while reducing the time required for matrix selection.
Solution Approach 2:
The patent performs preliminary classification of video blocks based on their characteristics (prediction mode, block size, gradient information) before the actual transform selection process. This preliminary action pre-filters the suitable transform matrices for each block type, so that during encoding, the system only needs to choose from a reduced set of candidate matrices, significantly reducing encoding time while preserving the ability to achieve high encoding precision when needed.
Data Source
AI summary
A video processing method is provided to include performing a conversation between a current video block of a video and a coded representation of the video. The conversion comprises selecting, for the current video block of a video, a transform set or a transform matrix to be used in an application of a secondary transform tool to the current video block based on a characteristic of the current video block and applying the selected transform set or transform matrix to the current video block. Using the secondary transform tool, during encoding, a forward secondary transform is applied to an output of a forward primary transform applied to a residual of the current video block prior to quantization, or during decoding, an inverse secondary transform is applied to an output of dequantization of the current video block before applying an inverse primary transform.