A fast stereo matching algorithm for real-time robot application

The main purpose of this paper is to introduce a fast stereo matching algorithm based on false epipolar constraint for real-time robot applications. False epipolar constraint can largely reduce the search region during the stereo matching process, and speed up the process when combining with other matching algorithms. In this paper, a fast area-based matching algorithm with this constraint is implemented to prove the efficiency of this constraint in our robot application, Furthermore, a new rule to choose the parameters of false epipolar constraint is obtained.