Multi-Hypothesis Motion Compensation for P Frame Prediction Accuracy
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Traditional P frame motion compensation methods in video encoding are limited by the accuracy of prediction blocks due to reliance on a single motion vector, leading to increased code rates when using two motion vectors for B frame bidirectional compensation.
Innovation Solution
A P frame-based multi-hypothesis motion compensation method that uses adjacent encoded image blocks to derive multiple motion vectors through joint estimation and weighted averaging, improving prediction block accuracy without increasing the code rate.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If a single motion vector is used in P frame motion compensation, then the code rate is kept low, but the accuracy of the prediction block is not very high
Solution Approach 1:
The motion compensation process is segmented into multiple hypotheses: a first motion vector from traditional motion estimation and a second motion vector from gradient-based prediction. These segmented motion vectors are independently calculated and then combined through weighted averaging to produce the final motion vector, thereby improving prediction accuracy without requiring a single complex motion model that would increase code rate
Solution Approach 2:
The patent applies partial action by using only the necessary components of multiple motion vectors. Instead of encoding all possible motion information, it selectively combines the first motion vector (from motion estimation) and the second motion vector (from gradient prediction) only when needed, using weighted averaging to achieve improved accuracy while maintaining reasonable code rate
2Measurement precision
If two motion vectors are introduced for B frame bidirectional motion compensation, then the prediction block accuracy is improved, but the code rate increases
Solution Approach 1:
The patent merges the advantages of multiple motion vectors through weighted averaging. Instead of encoding separate motion vectors for each hypothesis, it combines the first motion vector and second motion vector into a single final motion vector using weighted averaging, thereby achieving the prediction accuracy benefits of multiple hypotheses while transmitting only one motion vector, thus avoiding code rate increase
Solution Approach 2:
The patent changes the parameter representation by using weighted averaging to combine motion vectors. The weights are determined based on the reliability of each motion vector source, allowing the system to adaptively adjust the contribution of each hypothesis. This parameter transformation enables improved accuracy through multiple hypotheses while maintaining a single motion vector in the bitstream
Data Source
AI summary
A P frame-based multi-hypothesis motion compensation method includes: taking an encoded image block adjacent to a current image block as a reference image block and obtaining a first motion vector of the current image block by using a motion vector of the reference image block, the first motion vector pointing to a first prediction block; taking the first motion vector as a reference value and performing joint motion estimation on the current image block to obtain a second motion vector of the current image block, the second motion vector pointing to a second prediction block; and performing weighted averaging on the first prediction block and the second prediction block to obtain a final prediction block of the current image block. The method increases the accuracy of the obtained prediction block of the current image block without increasing the code rate.