基于视觉伺服的小型四旋翼无人机自主飞行控制研究进展

 在GPS信号较弱甚至失效的环境下，视觉伺服能够通过视觉信息控制自主飞行，因此近年来视觉伺服在自主飞行控制领域受到广泛关注。根据获取的图像信息不同，可将视觉伺服分为基于位置的视觉伺服和基于图像的视觉伺服。与基于图像的视觉伺服相比，基于位置的视觉伺服位姿估计稳定，可直观地在直角坐标空间定义机器人运动，符合机器人工作方式，且控制器设计简单，但控制精度受摄像机和机器人标定精度的影响，且计算量较大。对于小型四旋翼无人机自主飞行控制的应用研究中，视觉伺服的实时性、精确性和鲁棒性尚待提高，且小型四旋翼无人机的智能化不高，在室内室外模式转换及室内协同控制方面还有广阔的发展空间。

A review of achievements in small quadrotor autonomous flight control based on visual servoing

In the absence of GPS signal, visual servoing can make use of visual information to control quadrotor autonomous flight, so it has been widely used in the field of autonomous flight control in recent years. This article reviews the achievements of small quadrotor autonomous flight control based on visual servoing. Depending on the difference of image information, visual servoing can be divided into position-based visual servoing and image-based visual servoing. Compared with image-based visual servoing, pose estimation of positionbased visual servoing is stabile, and position-based visual servoing can define the robot movement in the rectangular coordinate space, and the controller design of position-based visual servoing is simple. But the control precision of position-based visual servoing is affected by camera and robot calibration precision, and calculation is large. It is pointed out that for further application of small quadrotor UAV autonomous flight control, the real-time accuracy and robustness of visual servoing need to be improved, and the intelligence of small quadrotor UAV requires upgrading. Also the research should be strengthened in aspects such as indoor/outdoor mode conversion and indoor coordinate control.