Multi-View Stereopsis Match Expand Filter Technique
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Multi-view stereopsis techniques face challenges with poor accuracy and infeasible initialization processes for various types of images, particularly in 3D image modeling.
Innovation Solution
A match, expand, and filter technique is employed to match features across multiple images, expand the sparse set of patches to a dense set, and filter out erroneous patches, optionally converting the reconstructed patches into 3D mesh models.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional multi-view stereopsis techniques are used, then 3D model reconstruction can be performed, but the accuracy is poor and initialization processes are infeasible for various types of images
Solution Approach 1:
The patent performs preliminary feature matching to identify corresponding patches across multiple views before performing 3D reconstruction. This preliminary action of matching features and establishing correspondences beforehand enables more accurate and feasible initialization for various image types, resolving the contradiction between measurement precision and ease of operation.
2Quantity of substance
If feature matching is performed across multiple images, then sparse patches can be obtained, but the patches are insufficient for complete 3D reconstruction
Solution Approach 1:
The patent segments the 3D reconstruction problem into multiple stages: first obtaining sparse patches through feature matching, then expanding to dense patches through systematic sampling and interpolation, and finally filtering to remove erroneous patches. This segmentation allows the system to achieve both quantity (dense patches) and reliability (filtered accurate patches) that neither sparse matching alone nor dense sampling alone could provide.
Solution Approach 2:
The patent implements a feedback mechanism where patches are iteratively expanded and then filtered based on consistency checks across multiple views. Erroneous patches are identified and removed through feedback from photometric consistency and geometric constraints, ensuring that the final dense set of patches maintains high reliability while achieving complete coverage for 3D reconstruction.
Data Source
AI summary
In accordance with one or more aspects of a match, expand, and filter technique for multi-view stereopsis, features across multiple images of an object are matched to obtain a sparse set of patches for the object. The sparse set of patches is expanded to obtain a dense set of patches for the object, and the dense set of patches is filtered to remove erroneous patches. Optionally, reconstructed patches can be converted into 3D mesh models.