Tile-Based Image Decoding Motion Vector List Exclusion
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
High Efficiency Video Coding (HEVC) schemes face challenges in efficiently decoding images, particularly when parallel processing is required across tiles, leading to increased processing overhead and circuitry complexity due to synchronization needs and mismatched prediction motion vectors between encoding and decoding sides.
Innovation Solution
An image decoding method that includes a constraint information decoding step, list generation for prediction motion vectors, and block decoding, where the prediction motion vector list excludes temporal motion vectors from blocks outside the target tile, ensuring appropriate decoding without referencing other tiles and reducing processing overhead.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If temporal motion vectors from blocks outside the target tile are included in the prediction motion vector list, then decoding flexibility is improved, but decoding reliability deteriorates due to mismatched prediction motion vectors between encoding and decoding sides
Solution Approach 1:
The patent segments the picture into multiple tiles and applies different motion vector list generation rules to each tile. Specifically, for blocks in a target tile, the prediction motion vector list excludes temporal motion vectors from blocks outside the target tile, while including temporal motion vectors from blocks within the same tile. This segmentation ensures that each tile's decoding process is self-contained and reliable, while maintaining flexibility within tile boundaries.
2Productivity
If reference to other tiles is allowed during decoding, then processing efficiency is improved, but device complexity increases due to synchronization requirements
Solution Approach 1:
The patent extracts and removes the problematic temporal motion vectors from blocks outside the target tile from the prediction motion vector list. By taking out these external references, the decoding process no longer requires complex synchronization mechanisms to handle cross-tile dependencies, thereby reducing device complexity while maintaining efficient processing within each tile.
3Measurement precision
If the prediction motion vector list includes all available candidates, then decoding accuracy is improved, but processing overhead increases
Solution Approach 1:
The patent applies local quality by tailoring the composition of the prediction motion vector list to the specific location and context of each block within the target tile. Instead of uniformly including all available motion vector candidates across the entire picture, the method selectively includes only those temporal motion vectors that originate from blocks within the same tile, optimizing the balance between accuracy and processing overhead for each local region.
Data Source
AI summary
An image decoding method includes a constraint information decoding step of decoding constraint information indicating prohibition of reference from a target tile, which is one of a plurality of tiles obtained by partitioning a picture, to another tile, and a list generation step of generating a prediction motion vector list including a plurality of prediction motion vector candidates and a plurality of prediction motion vector indices each associated with a corresponding one of the plurality of prediction motion vector candidates. In the list generating step, a prediction motion vector list not including a temporal motion vector of a block on a lower right side of the target block is generated.