Inter-prediction Mode Derivation for High-Resolution Image Coding
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing image compression techniques struggle to efficiently compress high-resolution images like HD and UHD, which requires improved inter prediction methods and motion information derivation for effective encoding and decoding.
Innovation Solution
The proposed method and apparatus enhance image coding efficiency by effectively determining an inter-prediction mode, deriving motion information, and employing weighted prediction for improved accuracy, while also adaptively signaling weight number information based on specific flags.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If traditional inter-prediction methods are used for high-resolution images, then compression is achieved, but coding efficiency is insufficient
Solution Approach 1:
The patent applies parameter changes by introducing weighted prediction parameters (weights and offsets) to modify the prediction process. Different weight values are assigned to different reference blocks based on their reliability, allowing the system to adaptively adjust prediction accuracy while maintaining coding efficiency for high-resolution images.
Solution Approach 2:
The patent implements dynamics through adaptive weight selection mechanisms where the prediction weights are not fixed but dynamically determined based on motion characteristics and reference block quality. This allows the system to optimize prediction accuracy for each specific coding scenario while maintaining overall coding efficiency.
2Measurement precision
If explicit weighted prediction is applied to improve prediction accuracy, then inter-prediction accuracy improves, but signaling overhead increases
Solution Approach 1:
The patent applies local quality by differentiating the treatment of different reference blocks through separate weight parameters for L0 and L1 reference picture lists. This allows precise control over prediction accuracy for each reference list while avoiding unnecessary signaling overhead for blocks where simple prediction suffices.
Solution Approach 2:
The patent implements partial action by selectively applying explicit weighted prediction only where needed rather than universally. The system can choose to apply weights only to certain prediction blocks or certain reference lists, reducing signaling overhead while maintaining accuracy where it matters most.
3Measurement precision
If motion information is derived using multiple merge candidates, then prediction accuracy improves, but computational complexity increases
Solution Approach 1:
The patent applies segmentation by dividing the motion information derivation process into distinct stages: generating multiple merge candidates, evaluating them, and selecting the best match. This structured approach allows the system to manage computational complexity by processing candidates in an organized sequence rather than simultaneously evaluating all possibilities.
Solution Approach 2:
The patent implements preliminary action by pre-generating and storing multiple merge candidates before the actual prediction process. This allows the system to have motion information ready in advance, reducing real-time computational complexity during encoding/decoding while maintaining accuracy through the pre-computed candidate pool.
Data Source
AI summary
An image decoding method and apparatus according to the present disclosure can determine an inter-prediction mode of a current block, derive motion information of the current block according to the determined inter-prediction mode, and obtain a prediction block of the current block, on the basis of the derived motion information.