A framework for controlling wheelchair motion by using gaze information

Users with severe motor ability are unable to control their wheelchair using standard joystick and hence an alternative control input is preferred. In this paper a method on how to enable the severe impairment user to control a wheelchair via gaze information is proposed. Since when using such an input, the navigation burden for the user is significantly increased, an assistive navigation platform is also proposed to reduce the user burden. Initially, user information is inferred using a camera and a bite-like switch. Then information from the environment is obtained using combination of laser and Kinect sensors. Eventually, both information from the environment and the user is analyzed to decide the final control operation that according to the user intention and safe from collision. Experimental results demonstrate the feasibility of the proposed approach