Vision-based Attitude Determination And Runway Recognition For UAV Landing

Attitude determination and the runway recognition are very important for the landing of the Unmanned Aerial Vehicle (UAV). In this paper, we do some research on how to segment the runway area out of the image and obtain the attitude of the UAV with the vision technology. Based on our analysis of the video record of a certain type of UAV's flight trial, we present two algorithms: the horizon detection algorithm and the runway segmentation algorithm. The result shows that the two algorithms can work well in the image processing.In order to obtain the attitude information of the UAV, we research the relations between horizon, runway boundaries and the attitude. From the horizon line, the pitch and roll angles can be found and the yaw angle can be calculated from the runway boundaries. All this is useful for the realization of the UAV autonomous landing.