Obstacle avoidance of an electrical wheelchair using a stereoscopic camera system

Abstract

There are many handicapped people who are disabled by mental or physical problems around the world. Some of them have difficulty with mobility. For this reason, a wheelchair is one of the most popular apparatus to aid the handicapped with mobility.

In recent years, intelligent wheelchairs have been developed using advanced technology such as laser sensors, ultrasonic sensors, stereoscopic cameras, and others. In this thesis, an electric wheelchair is controlled to avoid obstacles and go through free spaces using a stereoscopic camera system. The main advantage of this system is the detection of obstacles in 3 dimensional environment. In comparison other sensors such as laser and ultrasonic sensors, can just detect obstacles in 2D environments.

The left and right images are collected from stereoscopic camera, which are used

to create disparity maps. Therefore, 3D images are contributed. From the information obtained by the 3D images, the wheelchair can calculate to detect obstacles and free spaces. Then it is controlled to avoid obstacles and go through free spaces. Experimental results in the indoor areas at International University illustrate the effectiveness of using the stereoscopic camera system