Video Signal Motion Refinement With CABAC Context Reuse
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing video compression technologies face challenges in improving coding efficiency for CABAC context models, inter prediction, intra prediction, and adapting to non-square transform blocks and adaptive in-loop filtering.
Innovation Solution
The method involves adaptively initializing a CABAC context model using the state of a previous picture or reference picture, refining motion vectors, using unidirectional or bidirectional intra prediction, selectively scanning transform coefficients, and applying in-loop filtering to virtual blocks with different motion vectors.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Manufacturing precision
If video resolution and quality are improved to meet increasing demand for high-definition and ultra-high-definition videos, then video quality is improved, but transmission and storage costs increase due to increased data amount
Solution Approach 1:
The patent applies parameter changes by utilizing CABAC (Context-Adaptive Binary Arithmetic Coding) technology to dynamically adjust encoding parameters and context models during video compression. This allows the system to achieve higher compression ratios while maintaining video quality, effectively reducing the data amount for transmission and storage without sacrificing manufacturing precision in video quality
2Productivity
If CABAC context model coding efficiency is improved by using previous picture states, then coding efficiency is improved, but system complexity increases due to additional context management
Solution Approach 1:
The patent applies preliminary action by pre-initializing the CABAC context model using state information from previous pictures before encoding the current picture. This preparatory step allows the encoder to start with optimized context values that reflect temporal correlations, improving coding efficiency without requiring complex real-time context management during the actual encoding process
Solution Approach 2:
The patent applies copying by replicating and reusing context model states from previous pictures that have similar characteristics to the current picture. Instead of managing entirely new context models, the system copies proven context configurations from historical data, reducing the computational burden of context management while maintaining high coding efficiency
3Measurement precision
If motion vector refinement is applied to improve inter prediction accuracy, then prediction accuracy is improved, but computational complexity increases
Solution Approach 1:
The patent applies partial action by performing motion vector refinement selectively rather than uniformly across all blocks. The system identifies specific regions or blocks that benefit most from refined motion vectors and applies the refinement process only to those areas, achieving improved prediction accuracy where needed while avoiding the excessive computational complexity of full-frame refinement
4Manufacturing precision
If in-loop filtering is applied to blocks with different motion vectors to improve video quality, then video quality is improved, but processing time increases
Solution Approach 1:
The patent applies local quality by performing in-loop filtering selectively on specific block boundaries where motion vector differences are significant. Instead of applying filtering uniformly to all blocks, the system identifies local regions with high motion variation and applies filtering only to those areas, improving video quality at block boundaries while minimizing the overall processing time increase
Data Source
AI summary
A video encoding/decoding apparatus according to the present invention acquires motion vector refinement information, performs motion compensation on the basis of a motion vector of a current block, refines the motion vector of the current block using at least one or both of the motion vector refinement information and the output of the motion compensation, and performs motion compensation using the refined motion vector.