Adaptive Cross-Component Filtering for Chroma Motion Compensation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing video coding techniques, such as VVC, struggle to achieve superior coding efficiency beyond the current standards, particularly in handling inter blocks for chroma components during video compression, leading to suboptimal bit-rate usage and video quality.
Innovation Solution
Implement adaptive cross-component filtering using an adaptive cross-component filter to generate motion-compensated chroma samples based on reconstructed luma samples, enhancing the coding efficiency of inter blocks.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If conventional motion compensation is used for chroma components, then the encoding process is simple, but coding efficiency is insufficient
Solution Approach 1:
The patent introduces an adaptive cross-component filter as an intermediary processing step between motion compensation and chroma prediction. This filter uses luma sample gradients to adaptively adjust chroma prediction, mediating between the simple motion compensation process and the need for improved coding efficiency. The filter acts as a bridge that enhances chroma prediction accuracy without requiring complete redesign of the motion compensation framework.
Solution Approach 2:
The patent dynamically changes filtering parameters (filter strength, filter type) based on local image characteristics derived from luma gradients. By adapting filter parameters to local variations in the image content, the system achieves improved coding efficiency in regions that benefit from filtering while avoiding unnecessary complexity in regions where simple motion compensation suffices.
2Measurement precision
If adaptive cross-component filtering is applied, then chroma prediction accuracy improves, but computational complexity increases
Solution Approach 1:
The patent applies filtering with local quality by using luma gradient information to determine where and how strongly to apply cross-component filtering. Instead of uniformly filtering all chroma blocks, the system adapts the filtering strength and application based on local image characteristics, achieving higher prediction accuracy only where needed and reducing unnecessary computational power consumption in regions where simple prediction suffices.
3Loss of energy
If standard chroma prediction is used, then bit-rate consumption is higher, but encoding complexity is lower
Solution Approach 1:
The patent implements a feedback mechanism where luma gradient information feeds into the chroma prediction process. The gradient calculations from luma samples provide feedback about local image characteristics that guide the adaptive filtering of chroma components. This feedback loop enables the system to reduce bit-rate consumption by making more accurate predictions that require fewer bits to encode, while the feedback itself is derived from already-available luma data.
Data Source
AI summary
Methods for video decoding and encoding, apparatuses and non-transitory computer-readable storage media thereof are provided. In one method for video decoding, a decoder may obtain a plurality of reconstructed luma samples for a current inter coding block, obtain an adaptive cross-component filter, and obtain a filtered motion compensated chroma sample based on the adaptive cross-component filter and the plurality of reconstructed luma samples.