Ernani M. Fernandes

A FEATURE BASED MOTION ESTIMATION FOR VEHICLE GUIDANCE

Motion estimation is a key issue in the research for intelligent vehicles. In this paper, we implement a real-time motion estimation technique based on Kanade- Lucas-Tomasi technique. This algorithm is composed of two parts: feature (cor- ners) selection, and tracking within consecutive frames. Experimental results on image sequence of a outdoor scene prove the high accuracy and robustness of the algorithm.