基于单目仿人机器人的障碍物测距方法①

摘要

In this paper, a monocular vision method is proposed to measure the distance between the robot and the obstacle. Firstly it converts the RGB image which is obtained from the robot to HSV image, then detects and identifies the obstacle in the HSV color space. Secondly it calculates the distance between the robot and obstacle using pinhole imaging principle and geometric coordinate transformation. Finally the method is tested on the Darwin humanoid robot;the result from the experiment proves the effectiveness and feasibility of the method.%　　为了能够实时测量机器人与障碍物的距离，提出一种单目视觉的方法。首先将机器人获取的 RGB 图像转换为HSV图像，在HSV颜色空间对障碍物进行检测和识别，然后利用小孔成像原理和几何坐标变换，计算障碍物与机器人的距离。最后将该方法应用到实验室 Darwin 仿人机器人上，实验结果表明该方法具有可行性和有效性。