A stereo vision system for pedestrian navigation

In this paper we present the application of a stereo vision system for pedestrian navigation. Corner detection, stereo matching, triangulation, tracking, and robust ego-motion estimation are used to estimate incremental ego-motion of the stereo cameras with particular focus on implementation on pedestrians. A novel robust ego-motion estimation algorithm was utilized to eliminate outliers, which are independent moving features, mismatched features in the stereo matching step and incorrect assigned features in the tracking step. We also introduce a new method based on the knowledge of gait analysis to capture images at the same stage of walking cycle. This leads to less winding trajectory, which can be tracked without increasing order and computational cost of the tracker. The whole navigation process has been experimentally verified.