Motion Estimation via Multi-Resolution Sampling
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Hierarchical motion estimation in image signal compression faces challenges in accurately determining object motion with high computational complexity and memory requirements, particularly in limited bandwidth communication systems.
Innovation Solution
The method involves performing first-type sampling to generate sample source and reference blocks, determining matching regions by comparing pixel values, and calculating motion vectors through these regions, using techniques like hexagonal system sampling to reduce sampling operations and increase sampling frequency.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If hierarchical motion estimation is used to reduce computational complexity and memory requirements, then processing efficiency is improved, but motion estimation accuracy deteriorates
Solution Approach 1:
The patent divides motion estimation into two stages: first performing estimation on down-sampled blocks to obtain coarse motion vectors, then refining these vectors through detailed comparison in the original resolution image. This segmentation allows the system to benefit from both low-complexity initial estimation and high-accuracy final refinement.
Solution Approach 2:
The patent performs preliminary motion estimation using down-sampled images before processing the full-resolution images. The motion vectors obtained from the down-sampled images serve as initial estimates that guide subsequent detailed analysis, reducing the search space and computational load for the final accurate estimation.
2Productivity
If down-sampling is applied to reduce data volume for motion estimation, then processing speed is improved, but motion detection precision deteriorates
Solution Approach 1:
The patent segments the image processing into two resolutions: down-sampled images for rapid initial motion estimation and full-resolution images for precise verification and refinement. This multi-resolution approach maintains processing speed while recovering precision through sequential analysis at different detail levels.
Solution Approach 2:
The patent introduces a resolution dimension by performing motion estimation at both down-sampled and full-resolution levels. This dimensional approach allows the system to exploit the speed advantage of lower resolution while ultimately achieving high precision through higher resolution analysis, effectively combining both benefits across different resolution dimensions.
Data Source
AI summary
An image frame motion estimation device and image frame motion estimation method using the same include performing first sampling for generating a first sample source block by performing first-type sampling on pixels of a source block; performing second sampling for generating a first sample reference block by performing first-type sampling on pixels of a reference block; determining a first matching region by comparing pixel values of the first sample source block and the first sample reference block; and determining a second matching region corresponding to the source block by comparing pixel values of a plurality of regions adjacent to the first matching region and the source block. Where one pixel is sampled for each block constituted by N pixels (N is a natural number) in width and M pixels (M is a natural number) in height and the sampled pixels are projected in a horizontal direction, the first-type sampling causes at least two pixels to be sampled for every M pixels or where one pixel is sampled for each block constituted by N pixels (N is a natural number) in width and M pixels (M is a natural number) in height and the sampled pixels are projected in a vertical direction, the first-type sampling causes at least two pixels to be sampled for every N pixels.