论文部分内容阅读
针对微型空中机器人在室内环境下无法借助外部定位系统实现自主悬停的问题,提出一种基于单目视觉的自主悬停控制方法.采用一种四成分特征点描述符和一个多级筛选器进行特征点跟踪.根据单目视觉运动学估计机器人水平位置;根据低雷诺数下的空气阻力估计机器人飞行速度;结合位置和速度信息对机器人进行悬停控制.实验结果验证了该方法的有效性.
In order to solve the problem of autonomous hovering of miniature airborne robots using an external positioning system in an indoor environment, an autonomous hovering control method based on monocular vision is proposed, using a four-component feature descriptor and a multi-level filter Feature point tracking.According to the monocular visual kinematics, the robot’s horizontal position is estimated, the robot’s flying speed is estimated based on the air resistance at low Reynolds number, and the robot is hovered based on the position and velocity information.The experimental results verify the effectiveness of the method.