机械制造与自动化Issue(1)：198-201,4.
基于单目视觉的障碍物定位和测量
Obstacle Detection and Measurement Based on Monocular Vision
王振 1王化明1
作者信息
- 1. 南京航空航天大学，江苏 南京210016
- 折叠
摘要
Abstract
Obstacle detection and measurement is one of the key problems for the autonomous movement of intel igent mobile robot. This paper presents an algorithm based on the combination of the appearance and the contact edge which is used to detect and measure the obstacle on the flat surface using single visual sensor. In the algorithm the image is primary handled according to the ap￣pearance property of known ground surface and the obstacle position defected according to the edge property of the contact area be￣tween obstacle and ground, then the reference area on the obstacle is seleted and the image is handled again according to the ap￣pearance property of the obstacle to get the position of the obstacle in the image,final y the obstacle detection and measurement are finished according to the principle of visual sensor. The effectiveness and accuracy of the algorithm is tested and verified on a wheeled mobile robot,the error of detection algorithm is about 5 %,while the error of measurement algorithm is about 1.89 %.关键词
单目视觉/外观属性/高度测量Key words
monocular vision/appearance property/height measurement分类
信息技术与安全科学引用本文复制引用
王振,王化明..基于单目视觉的障碍物定位和测量[J].机械制造与自动化,2015,(1):198-201,4.
