Detection of Obstacle and Freespace in an Autonomous Wheelchair Using a Stereoscopic Camera System

Computer Vision has been a very new field developing rapidly in recent years. One of the main advantages in this development is that the “human vision” abilities such as electronically perceiving and understanding of images are provided to apply to autonomous vehicles or mobile robots. In this paper, detection of obstacle and freespace in an autonomous wheelchair using a stereoscopic camera system is introduced. The image data captured from the real word are obtained from a multiple camera system called the stereoscopic camera system. This paper will represent two dimensional (2D) and three dimensional (3D) maps calculated based on disparity map using geometric projection. From the 2D and 3D maps, one can detect obstacles and freespaces for the objective of autonomous wheelchair control. The experimental results obtained show to illustrate the advantage of the proposed method.