Motion Estimation via Lattice Block Sampling
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional motion estimation schemes for high definition images face increased computation complexity due to the large number of pixels and search regions, leading to significant time and effort requirements, especially when applied to ultra-high definition images.
Innovation Solution
A motion estimation system and method that samples blocks based on a lattice structure, using motion vectors of neighboring blocks to determine motion vectors for both sampled and unsampled blocks, thereby reducing computation and improving accuracy through interpolation.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional motion estimation scheme is applied to high definition images, then motion estimation can be performed for all blocks, but computation complexity increases significantly
Solution Approach 1:
The image frame is divided into multiple blocks, and only a subset of these blocks (sampled blocks) undergo full motion estimation while other blocks (unsampled blocks) use simplified methods. This segmentation allows the system to maintain motion estimation accuracy for critical blocks while reducing overall computation complexity by processing fewer blocks in full detail.
Solution Approach 2:
Instead of performing complete motion estimation on all blocks, the system performs partial motion estimation only on sampled blocks selected via lattice structure. This partial action approach reduces computation complexity while maintaining sufficient accuracy by focusing detailed processing only where necessary.
2Measurement precision
If conventional motion estimation is applied to ultra high definition images, then all pixels are processed, but processing time increases by more than 16 times
Solution Approach 1:
The ultra high definition image is segmented into blocks, with only sampled blocks undergoing full motion estimation. This reduces the number of pixels processed from all pixels to approximately 25% of blocks, dramatically reducing processing time while maintaining accuracy for the sampled regions.
Solution Approach 2:
The system performs preliminary sampling to identify which blocks require full motion estimation before proceeding with detailed processing. This preliminary action allows the system to avoid unnecessary full processing of all blocks, reducing overall processing time while ensuring accurate motion estimation where needed.
3Device complexity
If block sampling based on lattice structure is used, then computation complexity is reduced, but motion estimation accuracy may be compromised for unsampled blocks
Solution Approach 1:
Motion vectors from neighboring sampled blocks serve as intermediaries to estimate motion vectors for unsampled blocks. Instead of directly computing motion for all blocks, the system uses the motion information from sampled blocks as intermediate data to infer motion in unsampled regions, maintaining accuracy while reducing computation.
Solution Approach 2:
The system copies motion vector information from neighboring sampled blocks and applies it to unsampled blocks. This copying approach allows unsampled blocks to inherit motion estimation results from similar regions, maintaining reasonable accuracy without performing full motion estimation on every block.
Data Source
AI summary
A system and method of estimating a motion of an image using block sampling are provided. The motion estimation system includes: a block sampling unit to divide a frame of an input image into a plurality of blocks and sample the plurality of blocks based on a lattice structure; a first motion vector determination unit to determine a motion vector of a first current block using motion vectors of neighboring blocks of the first current block with respect to the sampled blocks; and a second motion vector determination unit to determine a motion vector of a second current block based on features of motion vectors of sampled neighboring blocks of the second current block wherein the second current block is not sampled.