Mesh Stitching for Occlusion Gaps in Stereo Motion Estimation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current techniques for motion estimation (ME) and depth from stereo (DFS) face difficulties in handling occlusion or disocclusion, resulting in poor quality vectors or disparity values due to content being entirely present or entirely not present in a frame pair, which bi-directional searches fail to address effectively.
Innovation Solution
A method involving mesh stitching is employed to detect regions with motion disparity values below a coherency threshold, calculate adjustments to the mesh, and output these adjustments to interpolate edge values efficiently, thereby addressing gaps in bi-directional searches.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If bi-directional searches are used for motion estimation and depth from stereo, then computational coverage is improved, but poor quality vectors or disparity values still occur in occlusion or disocclusion regions
Solution Approach 1:
The patent introduces mesh stitching as an intermediary technique that bridges occlusion or disocclusion regions by interpolating motion vectors and disparity values from surrounding valid regions. This mediator approach fills the gaps left by bi-directional searches without requiring complex specialized handling for each occlusion case.
Solution Approach 2:
The patent changes the parameter approach by using mesh-based interpolation with adjustable stitching parameters to combine results from bi-directional searches. This allows continuous adjustment of motion estimation quality in occlusion regions while maintaining computational efficiency.
2Manufacturing precision
If mesh stitching is applied to fill gaps in occlusion regions, then interpolation quality is improved, but computational overhead increases
Solution Approach 1:
The patent applies mesh stitching selectively only in occlusion or disocclusion regions rather than across the entire frame. This partial action approach maintains high interpolation precision where needed while avoiding unnecessary computational overhead in already-valid regions.
Solution Approach 2:
The patent segments the frame into valid regions and occlusion regions, applying different processing strategies to each. This segmentation allows efficient handling by focusing computational resources only on the problematic occlusion areas where mesh stitching is needed.
Data Source
AI summary
This disclosure provides systems, devices, apparatus, and methods, including computer programs encoded on storage media, for mesh stitching for motion estimation and depth from stereo. A processor may obtain an indication of a mesh associated with a set of motion vectors for a first frame and a second frame in a plurality of frames. The processor may detect a set of regions in the mesh including a set of motion disparity values associated with the set of motion vectors, where each motion disparity value in the set of motion disparity values is less than a coherency threshold. The processor may calculate an adjustment to the mesh based on the region. The processor may output an indication of the adjustment to the mesh.