Novel Method of Obstacle Detection and Avoidance for Visually I mpaired People

Abstract

Electronic mobility aid, transforms visual information to another sensory modality, has proved to be useful for visually impaired people to commute confidently and independently. With recent developments in technology, more visual information can be provided that can assist the user in the better way to avoid obstacles. The paper is focused on portable mobility aid prototype based on stereo imaging that can help user avoid collision with the obstacles. The algorithm is based on segmentation of disparity image to detect obstacles in each segment and identify probable path free of obstacles. The information about the free path is conveyed to the user with the help of two vibrotactile sensors. The prototype was tried on visually impaired users. The experiments were conducted in terms of detection of obstacles, avoidance of obstacles and walking speed of visually impaired user in closed area with different number of obstacles. We conclude that mobility aid prototype is potentially effective for visually impaired users.