基于机器视觉道路识别算法的研究  

Research of the road recognition algorithm based on the machine vision

作　　者：黄涛[1] 丁俊进[1] 

机构地区：[1]武汉理工大学信息工程学院,湖北武汉430070

出　　处：《国外电子元器件》2007年第4期34-38,共5页International Electronic Elements

基　　金：武汉市重点科技攻关项目(20033001005-5-3)

摘　　要：在道路图像中,为了得到较理想的车道标识边缘,采用基于Sobel算子边缘图像增强的方法增强道路车道线的边缘,从而明显提高车道线提取的准确性。针对道路边界的形状特征提出的二次曲线道路边界模型,实现了对车道标识线的实时跟踪。In the road picture, in order to get the edge of identification of more ideal lane, a method based on Sobel operator edge picture strengthens is introduced,thus the accuracy that the lane thread is drawn will be increased significantly.Conic section road border model put forward to the form characteristic of the border of road, the ones that have realized to identification line of lane which are followed in real time.

关 键 词：机器视觉 边界跟踪 道路识别 边缘增强 

分 类 号：TP391.4[自动化与计算机技术—计算机应用技术]