Weighted Prediction Video Encoding Using Regional Block Segmentation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing video encoding methods using weighted prediction in standards like H.264/AVC face accuracy issues due to fixed weights for reference blocks, leading to degradation in image prediction quality, especially for images with changing brightness.
Innovation Solution
A method that estimates multiple motion vectors for interprediction and divides the current block into regions, allowing independent weight assignment to reference blocks for each region, enabling flexible weighted prediction and improved accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If fixed weights are assigned to reference blocks in weighted prediction, then the encoding complexity is reduced, but the prediction accuracy degrades for images with brightness changes
Solution Approach 1:
The current block is divided into multiple regions, and different weight settings are applied to each region. This segmentation allows the patent to use simple fixed weights for some regions while applying adaptive weights to other regions, thereby resolving the contradiction between encoding complexity and prediction accuracy.
Solution Approach 2:
Different weight strategies are applied to different regions of the current block based on local characteristics. Regions with significant brightness changes use adaptive weight calculation, while other regions use fixed weights, optimizing the balance between complexity and accuracy locally.
2Measurement precision
If adaptive weight calculation is performed for each reference block, then the prediction quality improves, but the encoding complexity increases
Solution Approach 1:
The patent segments the current block into multiple regions and applies adaptive weight calculation only to specific regions where brightness changes are significant, rather than uniformly applying adaptive weights to the entire block. This reduces the overall encoding complexity while maintaining prediction quality where needed.
Solution Approach 2:
Instead of applying adaptive weight calculation to the entire current block, the patent applies it partially to specific regions where it provides the most benefit. This partial action approach reduces computational complexity while still improving prediction quality in critical areas.
3Reliability
If the current block is divided into multiple regions with independent weight assignment, then the prediction accuracy for brightness-changing images improves, but the device complexity increases
Solution Approach 1:
The patent divides the current block into multiple regions and assigns weights independently to each region. This segmentation enables accurate handling of brightness changes in different areas while maintaining a structured approach that manages device complexity through systematic processing.
Solution Approach 2:
The patent applies local quality by using region-specific weight assignment that adapts to local brightness characteristics. This allows high prediction accuracy in regions with brightness changes while using simpler weight schemes in other regions, balancing reliability and complexity.
Data Source
AI summary
Provided is a video encoding and decoding method and apparatus. The video encoding method includes selecting a plurality of reference blocks based on a plurality of motion vectors and setting weights assigned to the plurality of reference blocks independently for each of a plurality of regions of a current block in order to predict and encode the current block, thereby providing a new encoding mode allowing accurate prediction of the current block and thus improving the compression rate of video encoding.