A Game-Theoretic Approach to the Enforcement of Global Consistency in Multi-view Feature Matching

摘要

In this paper we introduce a robust matching technique that allows to operate a very accurate selection of corresponding feature points from multiple views. Robustness is achieved by enforcing global geometric consistency at an early stage of the matching process, without the need of ex-post verification through reprojection. Two forms of global consistency are proposed, but in both cases they are reduced to pairwise compatibilities making use of the size and orientation information provided by common feature descriptors. Then a game-theoretic approach is used to select a maximally consistent set of candidate matches, where highly compatible matches are enforced while incompatible correspondences are driven to extinction. The effectiveness of the approach in estimating camera parameters for bundle adjustment is assessed and compared with state-of-the-art techniques.