Image Decoding Motion Compensation Sub-Block Refinement
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional image encoding/decoding methods have limitations in improving coding efficiency due to their reliance on motion compensation in a block unit, which does not effectively adapt to the varying motion patterns within an image.
Innovation Solution
The proposed method refines motion information by deriving initial motion information from spatial and temporal neighbor blocks and pre-defined information, and then generates refined motion information using a bilateral template and distortion calculation, allowing for motion compensation in a sub-block unit.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If motion compensation is performed in a block unit using conventional methods, then the processing is simple and efficient, but the coding efficiency is limited due to inability to adapt to varying motion patterns within blocks
Solution Approach 1:
The patent divides the current block into multiple sub-blocks (first sub-block and second sub-block) and performs motion compensation independently on each sub-block using respective motion information. This segmentation allows the system to adapt to varying motion patterns within different regions of the block, thereby improving coding efficiency while maintaining manageable complexity through structured division.
Solution Approach 2:
The patent applies different motion information to different sub-blocks based on their specific motion characteristics. The first sub-block uses first motion information derived from first neighbor blocks, while the second sub-block uses second motion information derived from second neighbor blocks. This local quality approach enables optimized motion compensation for each region, improving overall coding efficiency without requiring complete redesign of the entire motion compensation process.
2Measurement precision
If motion information is refined using bilateral template and distortion calculation, then motion compensation accuracy is improved, but computational complexity increases
Solution Approach 1:
The patent performs motion information refinement selectively rather than universally. Refinement is applied to sub-blocks where it provides benefit, using bilateral templates and distortion calculations only when necessary to achieve accurate motion compensation. This partial action approach improves motion information accuracy for critical regions while avoiding unnecessary computational complexity across the entire image.
Solution Approach 2:
The patent uses distortion calculation as a feedback mechanism to evaluate and refine motion information. By comparing the bilateral template with the search area and calculating distortion values, the system iteratively improves motion information accuracy. This feedback loop enables precise motion compensation while managing computational complexity through targeted refinement based on distortion metrics.
3Adaptability or versatility
If conventional motion compensation is used without sub-block division, then the processing speed is faster, but the adaptation to different motion patterns is insufficient
Solution Approach 1:
The patent segments the current block into multiple sub-blocks, each processed with dedicated motion information. This segmentation enables the system to adapt to different motion patterns in different regions while maintaining processing efficiency through parallelization. The first sub-block and second sub-block can be processed simultaneously using their respective motion information, achieving both adaptability and speed.
Solution Approach 2:
The patent introduces dynamic motion information refinement where motion parameters are adjusted based on local characteristics. By deriving motion information from neighbor blocks and refining it through distortion calculation, the system dynamically adapts to varying motion patterns. This dynamic approach maintains processing speed while significantly improving adaptability to different motion scenarios.
Data Source
AI summary
The present invention relates to an image decoding method. The image decoding method comprises deriving initial motion information of a current block from at least one of motion information of a spatial neighbor block, motion information of a temporal neighbor block, and pre-defined motion information, generating refined motion information by performing motion information refinement for the initial motion information and generating a prediction block of the current block by using the refined motion information.