Development of Power Assist Wheelchair Control with Driving Assistance using Fish-eye Vision

Abstract

This paper presents a power assisting control system with driving assistance using monocular fish-eye vision for electric wheelchair. In our past researches, several approaches to vision-based power assist control have been studied in order to achieve driving assistance. In order to obtain wide angle view in front of wheelchair, the fish-eye vision system is employed, and the optical flow vectors are only used for simplified calculation process of driving assistance without complicated 3D reconstruction in our approach. Also, the noise influence of the optical flow vectors calculated by block matching algorithm is reduced for improving the practical performance. The validity of our system is evaluated by several experimental results.