Video Encoding Weight Prediction for Non-Rectangular Block Edges
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing video encoding technologies face challenges in accurately predicting and coding non-rectangular blocks, particularly at the edges of objects, leading to poor prediction effects and coding performance.
Innovation Solution
The proposed method involves configuring weight prediction angles and parameters for current blocks, determining surrounding matching positions, and calculating target and associated weight values to improve prediction accuracy and coding efficiency.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If rectangular partition is used for current block, then device complexity is reduced, but prediction accuracy deteriorates for non-rectangular object edges
Solution Approach 1:
The current block is divided into multiple triangular sub-blocks instead of using a single rectangular partition. This segmentation allows the prediction region to better conform to the actual object boundaries, improving prediction accuracy for non-rectangular shapes while maintaining manageable complexity through systematic triangular division
Solution Approach 2:
Different triangular sub-blocks within the current block are assigned different prediction modes and weight values based on their local characteristics. This allows each region to be predicted with optimal quality suited to its specific geometric and texture properties, rather than applying a uniform rectangular partition approach
2Measurement precision
If non-rectangular sub-blocks are used for partition, then prediction accuracy for object edges is improved, but coding performance deteriorates
Solution Approach 1:
Weight values are dynamically adjusted for different triangular sub-blocks based on their position and characteristics. By changing the weight parameters in the prediction formula, the system optimizes prediction accuracy for each region while maintaining a unified coding framework that preserves coding efficiency
Solution Approach 2:
The prediction system dynamically selects different prediction modes and weight configurations for different triangular sub-blocks based on their specific characteristics. This dynamic adaptation allows the system to achieve high prediction accuracy for complex object edges while maintaining overall coding performance through efficient mode selection
3Measurement precision
If weighted prediction with multiple parameters is implemented, then prediction accuracy is improved, but device complexity increases
Solution Approach 1:
Different weight values and prediction parameters are assigned to different triangular sub-blocks based on their local characteristics. This localized parameter assignment improves prediction accuracy for each region while the systematic organization of these local parameters prevents overwhelming complexity
Solution Approach 2:
The complex parameter configuration problem is segmented by dividing the current block into multiple triangular sub-blocks, each with its own optimized parameters. This segmentation transforms a single complex parameter set into multiple simpler, localized parameter sets that are easier to manage and optimize
Data Source
AI summary
Embodiments of the present disclosure provide an encoding and decoding method, apparatus and device. The method includes: obtaining a weight prediction angle and weight configuration parameters; configuring reference weight values for surrounding positions outside the current block; determining a surrounding matching position to which the pixel position points; determining a target weight value of the pixel position based on a reference weight value associated with the surrounding matching position, and determining an associated weight value of the pixel position based on the target weight value of the pixel position; determining a first prediction value of the pixel position and a second prediction value; determining a weighted prediction value of the pixel position based on the first prediction value, target weight value, second prediction value and associated weight value; determining weighted prediction values of the current block based on the weighted prediction value of each pixel position of the current block.