Inter-Frame Motion Vector Estimation Using Sub-Block Matching
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing image processing methods face challenges in accurately determining motion vectors during inter-frame coding, leading to large bit streams and reduced image quality due to suboptimal matching blocks.
Innovation Solution
The method involves downsampling and upsampling image frames, dividing them into sub-blocks, calculating similarities, and determining optimal matching blocks to obtain accurate motion vectors, thereby improving image quality and reducing bit stream size.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Quantity of substance
If inter-frame coding is performed with traditional matching block search, then the amount of encoded data is reduced, but the motion vector accuracy deteriorates leading to reduced image quality
Solution Approach 1:
The patent divides the image frame into multiple blocks and further subdivides each block into sub-blocks for motion estimation. This segmentation allows the system to search for matching blocks at multiple granularities, improving motion vector accuracy without proportionally increasing the amount of encoded data, as only the final selected motion vectors need to be transmitted.
Solution Approach 2:
The patent applies different matching criteria and search strategies to different blocks or regions of the image frame. By adapting the matching precision and method to local characteristics (e.g., high-motion vs. low-motion regions), the system achieves higher overall motion vector accuracy while encoding only essential data, thus resolving the contradiction between data amount and accuracy.
2Device complexity
If traditional matching block search is used, then encoding complexity is reduced, but matching block accuracy deteriorates resulting in large bit streams
Solution Approach 1:
The patent segments the motion estimation process into multiple stages: initial full-block matching, followed by sub-block refinement for selected regions. This segmented approach maintains manageable encoding complexity by avoiding exhaustive fine-grained search across the entire frame, while still achieving high matching accuracy in critical areas through targeted sub-block analysis.
Solution Approach 2:
The patent applies high-precision sub-block matching selectively to only those blocks that require it (e.g., blocks with high motion or at block boundaries), rather than applying exhaustive search to all blocks. This partial application of excessive precision maintains overall encoding efficiency while improving matching accuracy where it matters most, reducing the bit stream size.
3Speed
If simple matching criteria are used, then processing speed is improved, but motion estimation accuracy deteriorates leading to poor image quality
Solution Approach 1:
The patent implements a two-stage matching process: first using simple, fast criteria (e.g., sum of absolute differences at block level) to quickly identify candidate matching blocks, then applying more accurate but computationally intensive criteria (e.g., sub-block level comparison) only to these candidates. This segmentation of matching criteria maintains high processing speed while achieving accurate motion estimation.
Solution Approach 2:
The patent applies different matching criteria with varying precision to different regions or blocks based on their characteristics. High-precision matching is applied locally to critical regions (e.g., regions with complex motion or at block boundaries), while simpler criteria are used in other regions, thus maintaining overall processing speed while improving motion estimation accuracy where needed.
Data Source
AI summary
An image processing method includes: obtaining a current image frame and a reference image frame; sequentially performing downsampling and upsampling on the current image frame to obtain a processed current image frame, and sequentially performing downsampling and upsampling on the reference image frame to obtain a processed reference image frame; according to a preset division manner, dividing the processed current image frame into current image sub-blocks and dividing the processed reference image frame into reference image sub-blocks; determining a reference image sub-block with a minimum similarity to each current image sub-block among the reference image sub-blocks as a matching block of the current image sub-block; obtaining a motion vector corresponding to the current image sub-block based on each current image sub-block and the matching block corresponding to the current image sub-block; and encoding the current image frame based on the motion vector.