基于多帧叠加和窗口搜索的快速车道检测  

Efficient Method of Lane Detection Based on Multi-frame Blending and Windows Searching

作　　者：陈涵深[1,2] 姚明海 陈志浩[1] 杨圳 CHEN Han-shen;YAO Ming-hai;CHEN Zhi-hao;YANG Zhen(College of Information Engineering,Zhejiang University of Technology,Hangzhou 310023,China;Zhejiang Institute of Communications,Hangzhou 311112,China)

机构地区：[1]浙江工业大学信息工程学院,杭州310023 [2]浙江交通职业技术学院,杭州311112

出　　处：《计算机科学》2018年第10期255-260,共6页Computer Science

基　　金：浙江省教育厅项目(Y201635456);浙江省自然科学基金项目(LZ14F030001)资助

摘　　要：车道检测是辅助驾驶和自动驾驶的重要研究内容。针对现有车道检测算法的鲁棒性和复杂度较难均衡等问题,提出一种基于多帧叠加和窗口搜索的快速车道检测算法。首先,通过逆透视变换(IPM)把指定的感兴趣区域(ROI)转换成鸟瞰图,结合多帧叠加的方法把RGB图像转化成二值图。其次,根据近视场中的像素密度分布,计算当前帧的车道线起始点,并采用滑动窗口搜索的方法提取整个车道线。最后,根据车道线的特征,选择不同的车道模型,使用最小二乘法(LSE)拟合得到模型参数。大量的实际道路行驶测试结果表明,该算法能快速地检测车道线,并具有一定的鲁棒性和准确性。Lane detection is one of the most important research areas in assistance driving and automated driving.Many efficient lane detection algorithms have been proposed recently,but most of them are still hard to achieve a balance between computational efficiency and accuracy.This paper presented a real-time and robust approach for lane detection based on multi-frame blending and windows searching.Firstly,the image is cropped and mapped to create a bird’s-eye view of the road.Then,the RGB image is converted to a binary image based on a threshold of multi-frame blending.In the next step,the starting point of the lane line is calculated by using the pixel density distribution in the near field of view,and the whole lane is extracted by the method of sliding window search.Finally,according to the feature of candidate lane,different lane models are defined and chosen,and the model parameters are obtained by Least Square Estimation(LSE).The proposed algorithm shows good performance when tested on real-world data containing various lane conditions.

关 键 词：辅助驾驶 车道检测 多帧叠加 窗口搜索 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]