Block Vector Predictor Reordering for Video Coding Efficiency
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing video coding technologies face inefficiencies in compressing and decompressing video sequences due to redundant information, particularly in handling screen content and complex scenes, leading to suboptimal bitrate and quality trade-offs.
Innovation Solution
Adaptive reordering of block vector predictor candidates based on intra and inter prediction techniques, combined with advanced quadtree and multi-type tree partitioning, to optimize block partitioning and prediction modes for improved coding efficiency.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If block vector predictor candidates are processed in conventional order, then decoding complexity is reduced, but coding efficiency deteriorates due to suboptimal candidate selection
Solution Approach 1:
The patent implements dynamic reordering of block vector predictor candidates based on prediction mode type (intra vs inter prediction). The candidate list order is adaptively changed according to the prediction mode, allowing the system to optimize candidate selection for each specific mode while maintaining manageable decoding complexity through rule-based reordering rather than exhaustive search
Solution Approach 2:
The patent changes the ordering parameter of block vector predictor candidates based on prediction mode. For intra prediction modes, candidates are ordered differently than for inter prediction modes, optimizing the selection process for each mode type and improving coding efficiency without requiring complex adaptive algorithms
2Measurement precision
If more block vector predictor candidates are evaluated, then prediction accuracy improves, but computational complexity increases
Solution Approach 1:
The patent performs preliminary organization of block vector predictor candidates by prediction mode type before the actual prediction process. Candidates are pre-grouped and ordered according to whether they correspond to intra or inter prediction modes, allowing the decoder to efficiently select appropriate candidates without performing complex real-time analysis during decoding
Solution Approach 2:
The patent segments the block vector predictor candidate list into distinct groups based on prediction mode type. By separating candidates into intra-prediction groups and inter-prediction groups with different ordering, the system enables more accurate prediction while keeping the computational burden manageable through structured organization rather than exhaustive evaluation
Data Source
AI summary
A video coder (encoder or decoder) determines a sample, displaced from the current block by an amount indicated by a candidate block vector predictor (BVP), is outside a range for samples in a reference region, The reference region overlaps with a second reference region that defines valid positions of BVPs for the current block. The video coder constructs, based on the determining and the second reference region, a list of candidate BVPs for the current block. The list excludes the candidate BVP and comprises an adjusted candidate BVP where a second sample displaced from the current block by a second amount indicated by the adjusted candidate BVP has a coordinate inside the range of values. The video coder orders the list of candidate BVPs based on costs for the candidate BVPs in the list and codes the current block based on the ordered list of candidate BVPs.