Research On Rotorcrafts Indoor Navigation Using Vision/INS

In recent years,small Rotor Aerial Vehicle(RAV)is developing rapidly.Thanks to its high control accuracy and small size,the indoor environment application of rotorcraft UAV attracted more and more attention.However,due to limitations in indoor environment,the tradition navigation can't satisfy the requirements of the indoor navigation.In this paper,a new navigation method based on binocular vision is proposed.In this method,the output attitude information of vision navigation is corrected by integrating the output information of inertial navigation.Meanwhile,a fast stereo matching algorithm is designed for obstacle detection and the obstacle-avoidance path planning.The main contents of this paper are listed as follows:1.A stereo-vision navigation method is proposed.Through the research and analysis of Scale-invariant feature transform(SIFT)and oriented FAST and rotated BRIEF(ORB)algorithm,the algorithm of the extraction of the feature points is determined.The position and posture changes of UAV are calculated by the motion information of the camera,which are determined by the feature points.In addition,the cumulative errors in vision navigation is decreased by key frame method.2.A vision/INS integrated navigation method is proposed.The Kalman filter combing inertial navigation information and vision navigation information has been used to improve the accuracy of the navigation accuracy of UAV,and the parameters of the Kalman filter will be adaptively changed by the motion mode determined by the inertial sensor data.3.A fast stereo matching algorithm based on strong similarity is proposed,which is applied for UAV obstacle detection.A path planning algorithm for UAV is designed based on the A* heuristic algorithm.By analyzing the obstacle detection results and UAV speed,the map between nodes could be automatically generated and the path for avoiding the obstacles will be calculated.