Obstacle Detection Approach for Robotic Wheelchair Navigation

Abstract

Over the past decade, many efforts have been taken to give mobility to disabled and disordered persons. Many people are totally depending on their caregiver and their hospitality but caregiver cannot remain 24 h with them. These disabled and disordered people also want mobility and connections with surrounding environment and society. There is still need for the development of such devices which could control, monitor, maintain, and assist these people, so that they could lead their work and with greater ease. Augmentative technology has many issues to deal in a broad area dealing with deep R&D on assistive devices. Robotic wheelchair is a subarea of augmentative technology, which ensures safer and easier maneuverability of the user of the wheelchair both in indoor and outdoor environments, even in the absence of the caregiver. For safer movement in an unknown environment of a robotic wheelchair, designing of a proper obstacle detection system is important. This work deals with proposition of new approach for obstacle detection through image processing. The proposed system uses simple and logical approach to detect the obstacles in the path of robotic wheelchair. The approach has been tested on self-captured images with different backgrounds, target obstacles, and light conditions. The experimental result shows that the proposed approach can precisely detect the obstacles.