Research On Automatic Navigation Of Unmanned Aerial Vehicle Based On Computer Vision

With the rapid development of computer vision and micro-electronic technologies, it graduallybecomes a popular way to estimate the flight attitude and the relative position of Unmanned AerialVehicles (UAVs) via the image information in the field of automatic navigation, which provides a newmethod to the design of the high-level autonomous control systems.A navigation system based on computer vision is designed for the navigation in UAV Cable-hockrecovery approach, and this research proposes an automatic navigation algorithms based on computervision. An improved algorithm, based on template matching technique, is adopted to track targets thatchange drastically in size or appearance. The experiments show that the issue of navigationparameters acquired, and target recognized automatically as well as and stable real-time tracking issolved by the recovery system.This research proposed a movement estimating navigation strategy of UAV in the unknownscene. In the study of visual Simultaneous Localization And Mapping (SLAM) algorithm steps, theExtended Kalman Filter algorithm (EKF) predicts trajectory. Simulation experiments show that theUAV can be equipped with a camera to acquire the UAV position and attitude of navigationparameters, the trajectory of the real-time estimate, the sparse feature mapping and so on by theInertial Measurement Unit (IMU) and altimeter is solved with a certain theoretical and engineeringvalue.