Obstacle Detection for Unmanned Aerial Vehicle Movement by Using 3D Stereoscopic

Abstract

Obstacle is a hindrance to the unmanned aerial vehicle movement in various situations. The terrain has obstacle difficult movement effect, or the movement may be clash with obstacles. So the accuracy of the obstacle detection is very important for unmanned aerial vehicle movement. This paper proposes an obstacle detection system for unmanned aerial vehicle movement using correlation techniques of 3D Stereoscopic to examine the relationship of 3D images. The proposed system consists of two processes which are the obstacles detection in the movement vision, and objects avoidance by analysis of the correlation of 3D images. The
correlation can analyze an obstacle of the movement. The system performances are evaluated by using 3D stereoscopic images input to the system to analyze it. The experiments results showed that accuracy of the detected obstacles in the movement vision is 83.33 %, and objects avoidance in the movement vision is 85.71 %.