Vision-based Navigation and Obstacle Detection for UAV

Navigation and obstacle avoidance is essential for an unmanned air vehicle (UAV) to achieve fully autonomous flying through urban environment. For that purpose a new method for navigation and obstacle detecting is presented in this paper. In this method a camera is combined to a GPS/INS integrated navigation for improving navigation precision and detecting obstacle in the environment. A sequence of images is taken by the camera. From those images several feature points are extracted and matched. The direction of camera displacement is obtained by tracking these feature points. Meanwhile the calculated displacement of camera is computed using the output of GPS. So we can observe the error of displacement projection on the image plane to improving navigation precision. At the same time the obstacle information is gain from the sequence images.