摘要
随着无人机自主避障技术的发展,无人机的在线环境感知能力显得愈发重要;环境感知的重要组成部分便是障碍物的检测;与超声波、毫米波雷达这些只能检测出障碍物距离的传感器相比,视觉传感器具有独特的优势,因此成为多旋翼无人机进行自主避障的热点;采用双目立体视觉技术对多旋翼无人机前方视野内的环境进行障碍物检测,提出了一种基于柱状图的障碍物检测方法;首先,利用双目视觉传感器获取图像并进行图像处理;其次,把图像分成若干个柱状图的区域,提取每个柱状图区域里的障碍物深度信息;再次,检测柱状图区域里的障碍物相对于多旋翼无人机的方向;最后,通过室内环境进行算法验证;实验结果表明了该方法的有效性,并成功完成了室内未知环境下多旋翼前进方向的障碍物检测工作。
With the development of autonomous obstacle avoidance technology,the ability of UAV's(Unmanned Aerial Vehicle)online environment perception becomes more and more important.An important component of environmental perception is the detection of obstacles.Compared with ultrasonic and millimeter-wave radar,which can only detect the distance of obstacles,vision sensors have unique advantages,so they become the hot spots for autonomous obstacle avoidance of the UAV.This paper uses stereo vision technology to detect obstacles in the foreground of the UAV.In this paper,an obstacle detection method based on histogram is proposed.Firstly,the image is acquired by stereo vision sensor and image processing is performed.Secondly,the image is divided into several histogram regions,and the obstacle depth information is extracted in each histogram region.Thirdly,the obstacle in the histogram regions is relative to the direction of the UAV.Finally,the algorithm is verified by the indoor environment.The experimental results show that the method is effective,and the obstacle detection of forward direction is successfully completed in the unknown indoor environment.
作者
刘锐
陈凤翔
陈科羽
刘胜南
Liu Rui;Chen Fengxiang;Chen Keyu;Liu Shengnan(Guizhou Power Grid Co.Ltd., Guiyang, Guizhou 210046,China;Tianjin Zhong Wei Aerospace Data System Technology Co.Ltd., Tianjin Key Laboratory of Intelligent Information Processing in Remote Sensing, Tianjin 300301, China)
出处
《计算机测量与控制》
2018年第12期67-71,共5页
Computer Measurement &Control
