基于四线激光雷达的道路信息提取与目标检测
发布时间：2018-08-18 11:21
【摘要】：为了保障无人驾驶车在行驶过程中的安全性与可靠性,利用四线激光雷达对道路信息进行提取并对车辆前方的目标进行检测。针对传统的基于密度的DBSCAN算法对输入参数敏感、仅适用于单一密度数据集等缺点,提出了一种利用k-最近邻方法改进的DBSCAN算法,使算法参数(Eps,Minpts)可以根据数据集特点进行自适应的选取。并且根据激光雷达扫描到路沿上的数据集特点,提出基于共线点的二次提取算法,将路沿数据集准确的提取出来,并将车辆前方的道路划分为可行驶区域与不可行驶区域;在可行驶区域内,利用改进后的聚类算法检测道路中的障碍物。实车实验表明,本文所提出的算法稳定性强,在道路信息的提取与目标检测方面具有很好的实时性与准确性。
[Abstract]:In order to ensure the safety and reliability of the driverless vehicle in the course of driving, the road information is extracted by four-line lidar and the target in front of the vehicle is detected. Aiming at the disadvantages of traditional density-based DBSCAN algorithm which is sensitive to input parameters and only suitable for single density data set, an improved DBSCAN algorithm based on k- nearest neighbor method is proposed. The algorithm parameters (Eps-Minpts) can be selected adaptively according to the characteristics of the data set. According to the characteristics of the data set scanned to the road edge by lidar, a second extraction algorithm based on collinear points is proposed to extract the road edge data set accurately, and the road in front of the vehicle can be divided into driving area and non-driving area. The improved clustering algorithm is used to detect the obstacles in the driving region. Real vehicle experiments show that the proposed algorithm is stable and has good real-time and accuracy in road information extraction and target detection.
【作者单位】： 北京工业大学城市交通学院;
【基金】：北京市教委基金项目资助课题(JJ002790200802)
【分类号】：TN958.98;U463.6
[Abstract]:In order to ensure the safety and reliability of the driverless vehicle in the course of driving, the road information is extracted by four-line lidar and the target in front of the vehicle is detected. Aiming at the disadvantages of traditional density-based DBSCAN algorithm which is sensitive to input parameters and only suitable for single density data set, an improved DBSCAN algorithm based on k- nearest neighbor method is proposed. The algorithm parameters (Eps-Minpts) can be selected adaptively according to the characteristics of the data set. According to the characteristics of the data set scanned to the road edge by lidar, a second extraction algorithm based on collinear points is proposed to extract the road edge data set accurately, and the road in front of the vehicle can be divided into driving area and non-driving area. The improved clustering algorithm is used to detect the obstacles in the driving region. Real vehicle experiments show that the proposed algorithm is stable and has good real-time and accuracy in road information extraction and target detection.
【作者单位】： 北京工业大学城市交通学院;
【基金】：北京市教委基金项目资助课题(JJ002790200802)
【分类号】：TN958.98;U463.6
【相似文献】
相关期刊论文 前10条
1 袁方涛;孙侃;;道路信息发布预警系统应用及关键技术探讨[J];电子世界;2014年04期
2 ;美国将开发手机道路信息监测网[J];中国公路;2005年22期
3 陶静锋;吕植勇;张黎光;;多功能道路信息采集车控制系统设计[J];武汉理工大学学报(信息与管理工程版);2007年04期
4 程英伟;汤捷;;城市道路信息管网需求量预测分析[J];建筑经济;2009年06期
5 邓明君,熊坚,李贤林;驾驶模拟器道路信息数据库系统的开发与研究[J];昆明理工大学学报(理工版);2005年03期
6 徐兮;冯晓;;用边缘检测方法提取农村道路信息[J];四川测绘;2007年04期
7 黄亮;左小清;张晓晓;刘娅;;面向对象的道路信息识别提取分析[J];昆明理工大学学报(理工版);2010年06期
8 崔勇;王志良;孙e，
本文编号：2189312
本文链接：https://www.wllwen.com/kejilunwen/xinxigongchenglunwen/2189312.html
