Similarity and epipolar constraints for 3-d reconstruction

Abstract EN

Stereo matching is one of the most active research areas in computer vision.

In this paper we introduce a reconstruction approach to the local stereo matching problem using area-matching technique.

A large number of algorithms for area-matching have been developed [23], so we choose one of these algorithms to build our reconstruction approach.

The matching process is established between pixels in stereo pair using similarity and epipolar constraints.

In order to establish a common software platform and a collection of data sets for easy evaluation, we have designed a stand-alone, flexible Matlab implementation that enables the evaluation of the proposed approach and that can be easily extended include new algorithms.

The experiments on real images show the efficacy of the introduced approach to improve reconstruction accuracy.