Study And Realization On Indoor Autonomous Navigation System For Quad-copter

Unmanned Aerial Vehicle (UAV) navigation technology has been under great development in recent years. Nevertheless, most of the navigation technology is only applicable to the outdoor venues, thus there is an urgent need for indoor navigation technology research. Based on the flying platform of Quad-copter UAV, set up according to the indoor environment, the indoor autonomous navigation system is developed by the monocular vision assistant system and the laser rangefinder system. The major work and innovation of this paper are described as follow.Firstly, on the basis of comparison of several indoor navigation schemes and combination of the advantages of visual navigation and laser range finder navigation, both of the two schemes are utilized for indoor navigation. The "X" flying model is adopted in the implementation of indoor navigation through the analysis of the requirement of indoor flying environment. Indoor navigation system hardware, mounted on the quad-copter UAV, such as monocular camera, image transmission module, laser range finder and navigation processor is utilized to implement the indoor autonomous navigation. The Quad-copter UAV hardware system is built according to the selected scheme. Quad-copter UAV flight control software of attitude control loop, position control loop and the altitude control loop is researched, and the classic PID control is applied to control the quad-copter UAV.Secondly, assistant navigation of the monocular camera system is studied. Determining internal parameters of the camera through camera calibration, calculating pose of the window, according to the theory of PnP problem, with four concurrent planes features angular point of the window, calculating and distinguishing cipher text logo in accordance with the machine vision perspective deformable template matching method, the attitude angle of the cipher text logo relative to UAV is determined and the UAV is able to enter the target room consequently.Thirdly, creating local map with the data of laser rangefinder and constructing global map through data matching, thus self-locating of quad-copter UAV is implemented. According to the information provided by the monocular vision system for path planning, more intelligent indoor autonomous navigation is realized. When matching the laser rangefinder data, the improved ICP is utilized and the matching calculation time is reduced, therefore real-time of navigation is improved.Finally, experiments verify the feasibility of quad-copter UAV indoor navigation system and the mission of entering target room in complex environment with several corridors and rooms is achieved.