Picture Block Merge Mode Selection With Reduced Prediction Redundancy
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional video coding methods suffer from redundancy when determining merge modes for picture blocks, particularly when the available modes are limited, leading to inefficiencies in the prediction process.
Innovation Solution
A method is introduced to prioritize level-1 merge modes over level-2 modes and directly select level-2 modes like Triangle Partition Mode (TPM) or Combined Intra and Inter Prediction (CIIP) when level-1 modes are unavailable, reducing redundancy by bypassing the need to parse availability status information.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If conventional methods determine merge mode by checking availability of each candidate mode sequentially, then all possible merge modes can be considered, but redundancy occurs in the determination process
Solution Approach 1:
The patent segments the merge mode determination process into two distinct levels: level-1 merge modes (skip mode, merge mode, advanced motion vector prediction mode) and level-2 merge modes (triangle partition mode, combined intra and inter prediction mode). This segmentation allows the system to first check simpler level-1 modes, and only proceed to level-2 modes when level-1 is unavailable, thereby reducing redundant checks while maintaining comprehensive mode selection capability.
Solution Approach 2:
The patent establishes a predetermined priority order for checking merge modes, with level-1 modes checked before level-2 modes. This preliminary arrangement of checking order prevents redundant availability checks by ensuring that if level-1 modes are available, the system immediately selects from them without proceeding to check level-2 mode availability, thus eliminating unnecessary determination steps.
2Reliability
If all merge mode candidates are checked for availability, then the most suitable mode can be selected, but encoding and decoding efficiency decreases
Solution Approach 1:
By dividing merge modes into level-1 and level-2 categories with distinct availability checking requirements, the patent ensures that the most commonly used modes (level-1) are checked first. When level-1 modes are available, the system achieves high encoding efficiency without needing to check level-2 mode availability, while still maintaining the ability to select from all modes when needed.
Solution Approach 2:
The patent pre-establishes a hierarchy where level-1 merge modes take precedence over level-2 modes. This preliminary ordering allows the encoding and decoding processes to quickly determine availability and select modes without performing exhaustive checks on all modes, thereby improving processing efficiency while maintaining reliable mode selection through the structured availability determination.
Data Source
AI summary
Embodiments of this application disclose a picture prediction method and apparatus. The picture prediction method includes: determining whether a merge mode is used for a current picture block; when the merge mode is used for the current picture block, determining whether a level-1 merge mode is available for the current picture block; when the level-1 merge mode is unavailable, and a high-layer syntax element corresponding to a first merge mode indicates that the first merge mode is forbidden to be used, determining a second merge mode as a target merge mode used for the current picture block, where the first merge mode and the second merge mode are included in a level-2 merge mode; and predicting the current picture block based on the target merge mode. In the embodiments of this application, redundancy generated in a merge mode determining process can be reduced as much as possible.