视觉导航技术在小型无人机撞线回收中的应用

摘要

为了提高小型无人机回收的经济性和有效性，提出一种利用计算机视觉技术实现小型无人机撞线回收的导航系统。分析了撞线回收技术，对现有的目标实时跟踪法进行改进，通过采用基于模块匹配的改进算法实现对无人机目标的在线跟踪，使其能够对尺寸和外形都有较大变化的目标进行有效的在线跟踪，并通过仿真实现。实验结果表明：该跟踪算法跟踪速率可以达到15～25 Hz，能够实时解算其空间位置，准确地跟踪无人机目标，基本满足导航系统的实时性要求。%A navigation system based computer vision technology is adopted to meet efficiency and low cost requirement of small size UAV cable-hock recovery. It analyses the cable-hock recovery technology, and improves the existing real-time target tracking method by the use of improved algorithm based on template matching, so that it can tracking target online with a larger change on the size and shape in the simulation effectively. Experimental results show that:the tracking rate of the algorithm can reach 15~25Hz as well as get the real-time space position calculation, and track the UAV target accurately, which can basically meet the real-time requirements of navigation system.