OBSTACLE DETECTION AND AVOIDANCE USING WAVEFRONT PLANNER AND KINECT ON AWMRA

摘要

Obstacle avoidance in autonomous navigation platforms is a well known problem that can be solved in numerous ways. This paper considers and analyzes the use of wavefront planner as an obstacle avoidance algorithm for a 9-DoF wheelchair-mounted robotic arm (WMRA) [1], It also presents a suitable solution for obstacle detection using the OpenNI driver for interfacing with Microsoft's Kinect. It further analyzes the capabilities of an autonomous operation of the WMRA and explains how this algorithm can be implemented into its navigation control. The results of this project showed that the Kinect can provide a very accurate representation of the surroundings. The wavefront planner can use this data to find a path from a start position to a goal without running into an obstacle.