摘要
以基于计算机视觉技术实现四旋翼无人机自主着陆为目标,采用Pixhawk开源飞控和mavros技术,设计黑白相间的正方形分级嵌套地标,然后通过单目摄像头采集图像数据输入机载图像处理上位机NVIDIA TX2中进行图像处理,并结合OpenCV函数库功能完成地标检测识别,最终实现四旋翼无人机位姿解算与自主着陆.
This paper starts from the introduction to the navigation technology of the unmanned quad-rotor helicopter, takes the computer-based visual technology to realize the autonomous landing of the unmanned quad-rotor helicopter as the objective, designs a square cascade-nested landmark chequered with black and white by adopting the Pixhawk open-sourced flight control and the mavros technology, then collects image data and inputs into the airborne upper computer NVIDIA TX2 for image processing, and completes the landmark detection and identification relying on the OpenCV function library, thus realizing the pose position and orientation calculation as well as autonomous landing of the unmanned quad-rotor helicopter finally.
引文
[1] 岳基隆,张庆杰,朱华勇.微小型四旋翼无人机研究进展及关键技术浅析[J].电光与控制,2010,17(10):46-52.
[2] 方璇,钟伯成.四旋翼飞行器的研究与应用[J].上海工程大学学报,2015,29(2):113-118.
[3] 曹美会,鲜斌,张旭,等.基于视觉的四旋翼飞行器无人机自主定位与控制系统[J].信息与控制,2015,44(2):190-196.
[4] 毛臣健.基于多传感器信息融合的机器人自定位方法[J].科技通报,2012,28(12):146-148.
[5] 郑晓平.基于视觉的小型无人直升机自主降落导航系统的设计和研究[D].杭州:浙江大学,2013.
[6] OTSU N.A threshold selection method from gray-level histogram[J].IEEE Transactions System,Man,and Cybernetics,1979,9(1):62-66.
[7] 高翔,张涛.视觉SLAM十四讲:从理论到实践[M].北京:电子工业出版社,2017:85-103.
[8] ZHANG Z.A Flexible New Technique for Camera Calibration[J].IEEE Transactions on Pattern Analysis and Machine Intelligence,2000,22(11):1330-1335.