无人机视觉自主着陆仿真系统

计算机视觉导航技术具有精度高、不受电子干扰，成本低等特点，被誉为未来无人机自主着陆的必备手段之一。作为新一代着陆技术手段，视觉自主着陆技术需要进行大量的算法研究和飞行试验，因此有必要建立一个在实验室环境来验证所提方案的可行性。本文将虚拟现实、 可视化技术与计算视觉导航算法紧密结合起来，开发了一种无人机视觉自主着陆仿真验证系统， 构建场景模块，能够在复杂三维地形、不同气象条件下对无人机视觉自主着陆算法进行仿真验证， 实时计算和输出无人机着陆所需位置姿态参数。实验结果表明：该系统真实反映了无人飞行器飞行过程中的动态特性以及姿态角等的变化，并且具备良好的用户显控界面，验证了视觉自主着陆算法的可行性。

Simulation System of UAV Based Visual Autonomous Landing

Computer vision navigation technology will be one of the indispensable means for Unmanned Aerial Vehicles (UAVs) to take off and land autonomously due to their characteristics of high precision, not affected by electrical interference and low cost. As a new generation of navigation technology, Visual Autonomous Landing technology requires a lot of research algorithm and flight test .So it is necessary to establish a laboratory environment to verify that the scheme is feasible. In this paper, the virtual reality, visualization technology and computational visual navigation algorithm are closely combined to develop a visual autonomous landing simulation verification system for UAVs. The visual autonomous landing algorithm of UAVs in complicated 3D terrain and different weather conditions can be simulated and validated. This system can calculate and output the position and attitude parameters of UAV landing in real time .The experiment results show that: the system truly reflect the dynamic parameters of UAV during the flight and the change of attitudes, and also have a good user display and control interface. The experiment results verify the validity of vision Autonomous Landing algorithms.