基于激光雷达的道路边界检测的研究  

A study on a road edge detection based on laser radar

作　　者：谭宝成[1] 严征[1] 

机构地区：[1]西安工业大学电子信息工程学院,陕西西安710021

出　　处：《电子设计工程》2015年第17期80-82,86,共4页Electronic Design Engineering

基　　金：中央财政支持地方高校专项发展基金(CXY1080)

摘　　要：针对在自然环境下无人驾驶智能车辆的道路检测问题:本文提出了一种基于激光雷达的道路边界检测算法,该算法依据道路区域与非路区域存在的高度跳变的特性,利用边界是直线形状的特征,首先分别出道路区域和非道区域提取出边界点信息,然后把得到的边界点进行阈值处理和均值滤波,最后利用线性分析拟合得到道路的边界信息。试验结果表明,该边界检测算法可靠性强,稳定性高,能够准确完成道路边界检测任务,满足实时系统的要求。For The unmanned intelligent vehicle in the natural environment road detection problem：This paper presents a laser radar-based road boundary detection algorithm, the algorithm based on the highly hopping characteristics of the road area and non-road area, using the boundary is a straight line shape features, respectively, Firstly, respectively the roads area and non-road area extracted boundary points information, then the boundary points obtained by thresholding and mean filter, and finally analyzed using linear fitting road boundary information. Experimental results show that the boundary detection algorithm strong reliability, high stability, to complete road boundary detection tasks accurately and meet the requirements of real-time systems.

关 键 词：无人驾驶智能车辆 激光雷达 边界检测 均值滤波 线性分析 

分 类 号：TN98[电子电信—信息与通信工程]