Chroma Intra Block Copy Coding Without Explicit Block Vectors
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current video coding technologies face inefficiencies in intra picture block compensation, particularly when dealing with separate coding tree structures for luma and chroma components, which can lead to suboptimal compression ratios and increased computational complexity.
Innovation Solution
The method involves using separate coding tree structures for luma and chroma components, where the prediction mode flag determines the intra block copy mode, and block vectors are derived from co-located luma units to reconstruct chroma units, allowing for efficient intra picture block compensation without explicit signaling of block vectors for chroma units.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If separate coding tree structures are used for luma and chroma components, then coding efficiency is improved, but device complexity increases
Solution Approach 1:
The patent applies segmentation by separating the coding tree structures for luma and chroma components into independent trees. This allows each component to be coded separately with its own coding units and prediction units, enabling more flexible and efficient processing of each color component while maintaining the ability to handle their different characteristics optimally.
Solution Approach 2:
The patent implements a unified intra block copy prediction framework that works for both luma and chroma components. The same prediction mechanism, block vector derivation, and reference block selection processes are applied universally to both color components, simplifying the overall system design while accommodating separate coding trees through the multi-functional prediction unit.
2Manufacturing precision
If block vectors are explicitly signaled for chroma units, then reconstruction accuracy is improved, but bitrate increases
Solution Approach 1:
The patent uses copying by deriving chroma block vectors from co-located luma block vectors. Instead of explicitly signaling separate block vectors for chroma units, the system copies the block vector information from the luma component at the same spatial location, assuming that the motion or displacement patterns are similar for both color components. This significantly reduces bitrate while maintaining adequate reconstruction accuracy.
Solution Approach 2:
The patent merges the block vector signaling for luma and chroma components by using a single block vector for both color components at co-located positions. This combining approach reduces the total number of block vectors that need to be encoded and transmitted, thereby reducing bitrate while still enabling accurate reconstruction of both luma and chroma blocks through the unified intra block copy prediction.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
Aspects of the disclosure provide method and apparatus for video coding. In some examples, an apparatus includes receiving circuitry and processing circuitry. The processing circuitry decodes prediction information of a chroma coding unit in a segment (such as a slice, a tile, a group of tiles) of a current picture from a coded video bitstream. The segment lacks a temporal reference picture and having separate coding tree structures for luma and chroma components. Then, the processing circuitry detects a prediction mode flag associated with the chroma coding unit from the coded video bitstream, and determines, based on the prediction mode flag, a prediction mode from an intra prediction mode and an intra block copy mode for the chroma coding unit. Further, the processing circuitry reconstructs at least one sample of the chroma coding unit according to the determined prediction mode.