Video Coding Template Matching Fusion for Prediction Accuracy
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current video coding techniques are limited by the use of a single template pattern for template matching, leading to sub-optimal performance in predicting video blocks and resulting in inefficient bandwidth usage.
Innovation Solution
Implementing a video coder that stores multiple template patterns and constructs a template matching candidate list to generate a prediction block by fusing reference blocks based on these patterns, allowing for improved prediction accuracy and reduced residual information.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If a single template pattern is used for template matching, then the device complexity is reduced, but the prediction accuracy deteriorates
Solution Approach 1:
The patent divides the template matching process into multiple independent template patterns (e.g., L-shaped templates of different orientations and sizes). Each template pattern is evaluated separately to generate candidate prediction blocks, and the results are combined. This segmentation allows the system to capture different local characteristics of the video block without requiring a single complex template structure.
Solution Approach 2:
The patent merges the results from multiple template matching operations by combining the candidate prediction blocks generated from different template patterns. The merging process selects and integrates the most suitable prediction blocks to form the final prediction, thereby improving overall prediction accuracy while maintaining manageable complexity through systematic combination.
2Measurement precision
If multiple template patterns are used for template matching, then the prediction accuracy is improved, but the computational complexity increases
Solution Approach 1:
The patent applies partial action by selectively using multiple template patterns only when they provide meaningful improvements in prediction accuracy. Not all template patterns are applied uniformly to every block; instead, the system evaluates and applies templates based on local characteristics, reducing unnecessary computational overhead while maintaining accuracy benefits where needed.
Solution Approach 2:
The patent changes parameters such as template size, shape, and orientation to adapt to different local characteristics of video blocks. By varying template parameters dynamically rather than using a fixed template structure, the system achieves better prediction accuracy across diverse block types without requiring an excessive number of fixed template patterns.
3Loss of information
If multiple template patterns are used for template matching, then the residual information is reduced, but the bandwidth efficiency is improved only with increased processing
Solution Approach 1:
The patent performs preliminary template matching using multiple patterns during the encoding phase to generate highly accurate prediction blocks before residual calculation. By establishing accurate predictions in advance through multi-template evaluation, the system minimizes the residual information that needs to be encoded and transmitted, thereby reducing overall bitrate requirements despite the initial processing complexity.
Data Source
AI summary
A method of encoding or decoding video data includes constructing a template matching candidate list for a current block of the video data based on a plurality of template patterns; and encoding or decoding the current block based on the template matching candidate list.