基于多帧叠加和窗口搜索的快速车道检测

doi:10.11896／j.issn.1002-137X.2018.10.047

计算机科学 ›› 2018, Vol. 45 ›› Issue (10): 255-260.doi: 10.11896／j.issn.1002-137X.2018.10.047

基于多帧叠加和窗口搜索的快速车道检测

陈涵深^1,2, 姚明海¹, 陈志浩¹, 杨圳¹

浙江工业大学信息工程学院杭州310023 ¹
浙江交通职业技术学院杭州311112 ²

收稿日期:2017-08-12 出版日期:2018-11-05 发布日期:2018-11-05
作者简介:陈涵深(1983-),男,博士生,工程师,主要研究方向为计算机视觉、机器学习,E-mail:chs9811@163.com;姚明海(1963-),男,博士,教授,主要研究方向为机器学习、模式识别,E-mail:yhm@zjut.edu.cn(通信作者);陈志浩(1993-),男,硕士生,主要研究方向为计算机视觉、机器学习;杨圳(1994-),男,硕士生,主要研究方向为机器学习。
基金资助:
浙江省教育厅项目(Y201635456),浙江省自然科学基金项目(LZ14F030001)资助

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

CHEN Han-shen^1,2, YAO Ming-hai¹, CHEN Zhi-hao¹, YANG Zhen¹

College of Information Engineering,Zhejiang University of Technology,Hangzhou 310023,China ¹
Zhejiang Institute of Communications,Hangzhou 311112,China ²

Received:2017-08-12 Online:2018-11-05 Published:2018-11-05

摘要/Abstract

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。

关键词: 车道检测, 窗口搜索, 多帧叠加, 辅助驾驶

Abstract: 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.

Key words: ADAS, Lane detection, Multi-frame blending, Windows searching

中图分类号:

TP391

陈涵深, 姚明海, 陈志浩, 杨圳. 基于多帧叠加和窗口搜索的快速车道检测[J]. 计算机科学, 2018, 45(10): 255-260. https://doi.org/10.11896／j.issn.1002-137X.2018.10.047

CHEN Han-shen, YAO Ming-hai, CHEN Zhi-hao, YANG Zhen. Efficient Method of Lane Detection Based on Multi-frame Blending and Windows Searching[J]. Computer Science, 2018, 45(10): 255-260. https://doi.org/10.11896／j.issn.1002-137X.2018.10.047

参考文献

[1]GURGHIAN A,KODURI T,BAILUR S V,et al.DeepLanes:End-To-End Lane Position Estimation using Deep Neural Networks [C]∥Computer Vision & Pattern Recognition Workshops.2016:38-45.
[2]LI J,MEI X,PROKHOROV D.Deep Neural Network for Structural Prediction and Lane Detection in Traffic Scene [J].IEEE Transactions on Neural Networks & Learning Systems,2016,28(3):1-14. [3]CHEN C,SEFF A,KORNHAUSER A,et al.DeepDriving: Learning Affordance for Direct Perception in Autonomous Dri-ving [C]∥IEEE International Conference on Computer Vision.2015:2722-2730.
[4]JEONG P,NEDEVSCHI S.Efficient and robust classification method using combined feature vector for lane detection [J].Circuits and Systems for Video Technology,2005,15(4):528-537.
[5]GOPALAN R,HONG T,SHNEIER M,et al.A learning ap- proach towards detection and tracking of lane markings [J].Intelligent Transportation Systems,2012,13(3):1088-1098.
[6]SATZODA R K,SATHYANARAYANA S,SRIKANTHAN T, et al.Hierarchical additive hough transform for lane detection[J].Embedded Systems Letters,2010,2(2):23-26.
[7]CHEN Q,WANG H.real-time lane detection algorithm based on a hyperbola-pair model[C]∥Intelligent Vehicles Sympo-sium.Tokyo:IEEE,2006:510-515.
[8]WU B F,LIN C T,CHEN Y L.Dynamic calibration and occlusion handling algorithms for lane tracking [J].IEEE Transactions on Industrial Electronics,2009,56(5):1757-1773.
[9]BORKAR A,HAYES M,SMITH M T.Robust lane detection and tracking with ransac and Kalman filter [C]∥IEEE International Conference on Image Processing.2010:3225-3228.
[10]LEE M,JANG C,SUNWOO M.Probabilistic lane detection and lane tracking for autonomous vehicles using a cascade particle filter [J].Proceedings of the Institution of Mechanical Engineers Part D Journal of Automobile Engineering,2015,229(12):1-15.
[11]JU H Y,LEE S W,PARK S K,et al.A Robust Lane Detection Method Based on Vanishing Point Estimation Using the Relevance of Line Segments [J].IEEE Transactions on Intelligent Transportation Systems,2017,PP(99):1-13.
[12]WANG W,SHEN J,SHAO L.Consistent video saliency using local gradient flow optimization and global refinement [J].IEEE Transactions on Image Processing,2015,24(11):4185-4196.
[13]WANG W,SHEN J,PORIKLI F.Saliency-aware geodesic video object segmentation [C]∥Computer Vision & Pattern.IEEE Recognition.2015:3395-3402.
[14]ALY M.Real time detection of lane markers in urban streets[C]∥ IEEE Intelligent Vehicles Symposium.2008:7-12.
[15]BERTOZZI M,BROGGI A.GOLD:A parallel real-time stereo vision system for generic obstacle and lane detection [J].IEEE Transactions on Image Processing,1998,7(1):62-81.
[16]FISCHLER M A,BOLLES R C.Random sample consensus:A paradigm for model fitting with applications to image analysis and automated cartography[J].Communications of ACM,1981,24(6):381-395.
[17]HAN G F,LI X M,WU X.Research of lane line detection in the vision navigation of unmanned vehicle[J].Fire Control Command Control,2015,40(6):152-158.(in Chinese)
韩广飞,李晓明,武潇.无人驾驶汽车视觉导航中车道线检测的研究[J].火力与指挥控制,2015,40(6):152-158.
[18]LI C,LIU H Z,YUAN J Z,et al.Real-time Lane Detection Algorithm Based on Inter-frame Correlation[J].Computer Science,2017,44(2):317-323.(in Chinese)
李超,刘宏哲,袁家政,等.一种基于帧间关联的实时车道线检测算法 [J].计算机科学,2017,44(2):317-323.
[19]CHEN W W,JIANG Y T,TAN D K.A Fast Lane Marking Recognition Algorithm Based on Edge Projection[J].Automotive Engineering,2017,39(3):357-363.(in Chinese)
陈无畏,蒋玉亭,谈东奎.一种基于边缘点投影的车道线快速识别算法[J].汽车工程,2017,39(3):357-363.
[20]SONG R,CHEN H,XIAO Z G,et al.Lane detection algorithm based on geometric moment sampling[J].Computer Science,2017,44(2):455-467.(in Chinese)
宋锐,陈辉,肖志光,等.基于几何矩采样的车道线检测算法 [J].中国科学:信息科学,2017,47(4):455-467.
[21]HALOI M,JAYAGOPI D B.A robust lane detection and depar- ture warning system[C]∥IEEE Intelligent Vehicles Symposium(IV).2015:126-131.

Metrics

Viewed

Full text

Abstract

Cited

Shared

Discussed

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

PDF (PC)

摘要/Abstract

引用本文

使用本文

参考文献

相关文章 3

Metrics

本文评价

推荐阅读 0

[1]	张琳娜,陈建强,陈晓玲,岑翼刚,阚世超. 面向行车视频目标实时检测的轻量级SSD网络 Lightweight SSD Network for Real-time Object Detection in Automotive Videos 计算机科学, 2019, 46(7): 233-237. https://doi.org/10.11896/j.issn.1002-137X.2019.07.035
[2]	李堃, 黎向锋. 基于日间行车的灯语识别技术 Lamp Language Recognition Technology Based on Daytime Driving 计算机科学, 2019, 46(11A): 277-282.
[3]	李星,郭晓松,郭君斌. 基于HOG特征和SVM的前向车辆识别方法 HOG-Feature and SVM Based Method for Forward Vehicle Recognition 计算机科学, 2013, 40(Z11): 329-332.