Research On Target Tracking Of Multi Rotor Aircraft Based On Image Features

With the continuous development of modern science and technology,the technology of multi rotor aircraft is becoming more and more mature.In particular,it is more widely used in the application field.Image based UAV recognition and tracking technology is still a difficult problem to be solved at present.In this dissertation,six rotor aircraft and image acquisition system are used as experimental platform,and the target recognition and tracking problem of multi rotor aircraft based on image features is discussed in detail.The hardware and software composition of the system are introduced.The attitude control algorithm of multi rotor aircraft and the target image matching and tracking algorithm are emphatically analyzed,and the control method and the system work process of the target tracking system based on the image feature are summarized.The target tracking technology of multi rotor aircraft based on image features is studied in this dissertation,including the optimization of the attitude control algorithm of multi rotor aircraft,the matching of target image feature recognition,the hardware system and software system of multi rotor aircraft target tracking control and image acquisition target recognition and tracking.For multi rotor aircraft attitude control algorithm,based on the traditional PID algorithm,analyzes the principle of aerodynamics of the aircraft by quaternion and the number of aircraft attitude Euler solution,put forward a kind of position and attitude of double loop PID control optimization algorithm.The experimental results show that the improved multi rotor aircraft attitude control method is reliable,fast response and no static error and smaller overshoot,overcome the real-time vehicle,the disadvantages of slow response,and can resist certain interference and can quickly restore good performance stable equilibrium state.For target image recognition and tracking technology,based on the traditional region matching and feature matching method,a SURF image matching method combining region and feature information is proposed.The experimental results show that the improved matching method is not only small in computation,but also with low mismatch rate and strong robustness.Finally,the hardware platform of the multi rotor aircraft control system and the hardware platform of the image acquisition system are designed.The corresponding image acquisition MATLAB,target recognition and tracking algorithm,STM32 programming and optimization of vehicle attitude control are completed.The experimental results of target image recognition and tracking by multi rotor aircraft show that the design method of this dissertation can get satisfactory results.