Motion Estimation Circuit Using Scaled Frames for Vector Consistency
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing motion estimation methods for image processing result in inconsistent motion vectors for blocks within the same object moving at the same speed due to limited candidate blocks and image content variation, affecting motion compensation effectiveness.
Innovation Solution
An image processing method involving scaling down the frame to generate a scaled-down frame, performing motion estimation on both the original and scaled frames, and determining final motion vectors based on the original and adjusted vectors considering image content complexity.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If motion estimation is performed on each block independently using limited candidate blocks, then the calculation complexity is reduced, but the motion vector consistency within the same object deteriorates
Solution Approach 1:
The frame is divided into multiple blocks for independent motion estimation, while the scaling operation segments the problem into two levels: coarse motion estimation on scaled-down frame and fine adjustment on original blocks. This segmentation allows both computational efficiency and motion vector consistency.
Solution Approach 2:
The patent introduces a scaling dimension by creating a scaled-down version of the frame. Motion estimation is first performed on this reduced-dimensional representation, then the results are used to guide the motion estimation of blocks in the original frame. This dimensional transformation enables consistent motion vectors while reducing calculation complexity.
2Measurement precision
If multiple candidate blocks are used for motion estimation, then the motion vector accuracy for individual blocks is improved, but the computational load increases
Solution Approach 1:
Motion estimation is performed preliminarily on the scaled-down frame before processing the original frame. This preliminary motion vector from the scaled frame serves as a reference or initial value for the motion estimation of blocks in the original frame, reducing the search space and computational load while maintaining accuracy.
Solution Approach 2:
Instead of performing full motion estimation with extensive candidate blocks on all blocks in the original frame, the patent applies motion estimation partially: first on the scaled-down frame, then uses those results to guide the estimation on original blocks. This partial approach reduces computational load while achieving sufficient accuracy.
Data Source
AI summary
An image processing method includes the steps of: receiving an image signal, wherein the image signal comprises a frame; performing a motion estimation operation on a plurality of blocks within the frame to generate a plurality of first motion vectors, respectively; scaling down the frame to generate a scaled-down frame; performing the motion estimation operation on a specific block within the scaled-down frame to generate a second motion vector corresponding to the specific block; and determining a plurality of final motion vectors of the plurality of blocks of the frame according to the plurality of first motion vectors and the second motion vector.