Residue Transform for Multi-Channel Video Compression
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing image and video compression systems fail to efficiently exploit redundancy across different channels in luminance/chrominance color formats, leading to suboptimal compression efficiency.
Innovation Solution
The implementation of a residue transform process that applies integer operations across input channels to generate residue transformed data, reducing redundancy and enhancing compression efficiency by processing residual signals from multiple channels.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of manufacture
If compression is applied to individual luminance and chrominance channels separately, then the processing is simple and existing compression systems can be used, but redundancy among different channels is not exploited and compression efficiency is suboptimal
Solution Approach 1:
The patent combines multiple channels (luminance and chrominance) into a single processed stream by generating residue data representing differences between channels, then applying a single transform and quantization process to all channels together, rather than processing each channel independently through separate transforms and quantizers
Solution Approach 2:
The patent introduces a new processing dimension by creating residue data that represents inter-channel differences, then applying a transform that operates across channels rather than only within individual channels, effectively adding a cross-channel dimension to the compression process
2Productivity
If luminance/chrominance color format is used, then some redundancy among channels is removed, but substantial correlation among different channels remains and compression efficiency can be further improved
Solution Approach 1:
The patent performs preliminary processing by generating residue data that represents differences between channels before applying the transform, effectively pre-processing the data to highlight inter-channel correlations in a form that is more amenable to efficient compression
Solution Approach 2:
The patent changes the representation parameters by transforming residue data into a different domain using a custom transform matrix, converting the data into transform coefficients that have different statistical properties and are more suitable for compression
Data Source
AI summary
A method and apparatus for a multiple-channel image/video coding system are disclosed. A residue generation process is applied to the image/video data to generate residue data. A set of integer operations is applied to the residue data across the input channels to generate residue transformed data having multiple output channels. In one embodiment, the residue transformed data associated with a first output channel is related to the difference between a first residue data associated with a first input channel and a second residue data associated with a second input channel. In another embodiment, the residue transformed data associated with a second output channel is related to the second difference between a threshold and a third residue data associated with a third input channel, and wherein the threshold corresponds to the first truncated or rounded average of the first residue data and the second residue data.