Video Motion Compensation Using Sub-Block Optical Flow
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The increasing demand for high-resolution and high-quality images leads to higher data volumes, increasing costs in data transmission and storage, and existing video compression techniques are inefficient for stereographic content.
Innovation Solution
The method and apparatus apply bi-directional optical flow and overlapped block motion compensation on sub-blocks during video encoding/decoding, updating motion vectors and using motion refinement vectors based on average values of samples within the sub-blocks.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional image compression techniques are used for high-resolution video, then data transmission and storage costs increase, but image quality and resolution requirements cannot be met
Solution Approach 1:
The current block is divided into multiple sub-blocks for independent motion compensation processing. Each sub-block undergoes separate optical flow calculation and motion refinement, enabling more precise local motion representation while maintaining overall compression efficiency for high-resolution video
Solution Approach 2:
The patent applies bi-directional optical flow and motion refinement vectors to update motion vectors at sub-block level. These parameter changes enable more accurate motion compensation for high-resolution content, reducing the residual error and improving compression efficiency without sacrificing image quality
2Productivity
If bi-directional optical flow and motion refinement are applied on sub-blocks, then inter prediction efficiency improves, but computational complexity increases
Solution Approach 1:
By segmenting the current block into sub-blocks and applying motion refinement selectively, the patent achieves better prediction accuracy where needed while limiting the overall computational burden. The segmentation allows parallel processing of multiple smaller blocks rather than one large block
Solution Approach 2:
Motion refinement vectors are applied locally to specific sub-blocks based on their motion characteristics. This local quality approach ensures that computationally intensive operations are performed only where necessary to improve prediction accuracy, rather than uniformly across the entire block
3Measurement precision
If overlapped block motion compensation is applied on the current block, then prediction accuracy improves, but memory usage increases
Solution Approach 1:
Overlapped block motion compensation is applied at the sub-block level rather than the full block level. This segmentation reduces the amount of overlapping data that needs to be stored in memory, as each sub-block has smaller overlap regions with its neighbors
Solution Approach 2:
The patent uses motion refinement vectors to adjust motion parameters at the sub-block level, enabling more accurate prediction with smaller block sizes. This parameter change allows overlapped compensation to be applied more efficiently with reduced memory requirements compared to traditional full-block approaches
Data Source
AI summary
An image decoding method according to the present invention may comprise obtaining a motion vector of a current block, updating the motion vector when bi-directional optical flow is applied to the current block, performing motion compensation on the current block by using the updated motion vector.