首页 > 论文 > 中国激光 > 46卷 > 8期(pp:804002--1)

基于深度学习的道路标线自动提取与分类方法

Automatic Extraction and Classification of Road Markings Based on Deep Learning

  • 摘要
  • 论文信息
  • 参考文献
  • 被引情况
  • PDF全文
分享:

摘要

道路标线提取与分类是智慧城市建设中需要解决的关键技术之一,也是智能驾驶亟待解决的技术难题。提出了一种基于深度学习的道路标线自动提取与分类方法,通过移动窗口法结合相邻扫描线拓扑关系进行地面点云的提取,并生成强度图像,基于深度学习方法实现道路标线的自动提取与分类,并利用KD树聚类分割算法结合矢量化方案实现道路标线的矢量化。基于实验数据对该方法进行验证分析,结果表明,使用该方法进行自动提取与分类的精度和Fscore分别为92.59%和90.15%,证明了该方法的可行性和准确性。该方法为道路标线的自动提取提供了新思路,使道路标线提取工作变得更准确、高效,提升了道路标线获取与分类的智能化程度。

Abstract

Extraction and classification of road markings are two key technologies to be solved in the construction of an intelligent city and urgent technical problems that must be solved for intelligent driving. Therefore, herein, we propose a method of automatic extraction and classification for road markings based on deep learning. First, the ground point clouds are extracted through the moving-window method combined with the topological relations of adjacent scanning lines, and then the intensity images are generated. Automatic road-marking extraction and classification are realized based on the deep learning method. Road-marking vectorization is performed using the KD tree clustering algorithm and the vectorization scheme. The proposed method is analyzed based on the obtained experimental data. Results show that the precision and Fscore of the automatic road-marking extraction and classification reach 92.59% and 90.15%, respectively, proving the feasibility and accuracy of this method. Thus, the proposed method provides a new idea for automatic road-marking extraction and improves its accuracy, efficiency, and intelligent degree of road-marking acquisition and classification.

Newport宣传-MKS新实验室计划
补充资料

DOI:10.3788/CJL201946.0804002

所属栏目:测量与计量

基金项目:国家重点研发计划、高分辨率对地观测系统重大专项;

收稿日期:2019-02-26

修改稿日期:2019-04-02

网络出版日期:2019-08-01

作者单位    点击查看

黄刚:首都师范大学资源环境与旅游学院, 北京 100048北京四维远见信息技术有限公司, 北京 100070
刘先林:中国测绘科学研究院, 北京 100830

联系人作者:黄刚(hgminisar@163.com)

备注:国家重点研发计划、高分辨率对地观测系统重大专项;

【1】Peng J F. Study on automated extraction methods of highway elements from mobile lidar point clouds. Beijing: Beiing University of Civil Engineering and Architecture. 38-41(2017).
彭江帆. 基于车载激光扫描数据的高速公路道路要素提取方法研究. 北京: 北京建筑大学. 38-41(2017).

【2】Yan L and Li Z. Road marking extraction from mobile LiDAR data. Remote Sensing Information. 33(1), 1-6(2018).
闫利, 李赞. 车载激光点云道路标线提取方法. 遥感信息. 33(1), 1-6(2018).

【3】Li C, Creusen I, Hazelhoff L et al. Detection and recognition of road markings in panoramic images. ∥Jawahar C, Shan S. Computer vision-ACCV 2014 workshops. Lecture notes in computer science Cham: Springer. 9009, 448-458(2015).

【4】Foucher P, Sebsadji Y, Tarel J P et al. Detection and recognition of urban road markings using images. [C]∥2011 14th International IEEE Conference on Intelligent Transportation Systems (ITSC), October 5-7, 2011, Washington, DC, USA. New York: IEEE. 12385335, (2011).

【5】McCall J C and Trivedi M M. Video-based lane estimation and tracking for driver assistance: survey, system, and evaluation. IEEE Transactions on Intelligent Transportation Systems. 7(1), 20-37(2006).

【6】Yang M M, Wan Y C, Liu X L et al. Rapid automatic location and extraction for manhole covers based on ground mobile mapping system. Chinese Journal of Lasers. 45(8), (2018).
杨蒙蒙, 万幼川, 刘先林 等. 基于地面移动测量系统的井盖快速自动定位与提取方法的研究. 中国激光. 45(8), (2018).

【7】Guan H Y, Li J, Yu Y T et al. Using mobile laser scanning data for automated extraction of road markings. ISPRS Journal of Photogrammetry and Remote Sensing. 87, 93-107(2014).

【8】Kumar P. McElhinney C P, Lewis P, et al. Automated road markings extraction from mobile laser scanning data. International Journal of Applied Earth Observation and Geoinformation. 32, 125-137(2014).

【9】Holgado-Barco A, Riveiro B, González-Aguilera D et al. Automatic inventory of road cross-sections from mobile laser scanning system. Computer-Aided Civil and Infrastructure Engineering. 32(1), 3-17(2017).

【10】Chen G B, Gao Z H and He L. Step-by-step automatic calibration algorithm for exterior parameters of 3D lidar mounted on vehicle. Chinese Journal of Lasers. 44(10), (2017).
陈贵宾, 高振海, 何磊. 车载三维激光雷达外参数的分步自动标定算法. 中国激光. 44(10), (2017).

【11】Fang L N. Automated extraction of road geometric features from mobile laser scanning point clouds. Wuhan: Wuhan University. 52-66(2014).
方莉娜. 车载激光点云中道路环境几何特征提取. 武汉: 武汉大学. 52-66(2014).

【12】Yan L, Liu H, Tan J X et al. Scan line based road marking extraction from mobile LiDAR point clouds. Sensors. 16(6), (2016).

【13】Soilán M, Riveiro B, Martínez-Sánchez J et al. Segmentation and classification of road markings using MLS data. ISPRS Journal of Photogrammetry and Remote Sensing. 123, 94-103(2017).

【14】Deep learning. 1-15(2017).
Goodfellow I, Bengio Y, Courville A, Ian Goodfellow, Yoshua Bengio and Aaron Courville. 深度学习. 1-15(2017).

【15】LeCun Y, Bengio Y and Hinton G. Deep learning. Nature. 521(7553), 436-444(2015).

【16】Hu F, Xia G S, Hu J W et al. Transferring deep convolutional neural networks for the scene classification of high-resolution remote sensing imagery. Remote Sensing. 7(11), 14680-14707(2015).

【17】Wen C L, Sun X T, Li J et al. A deep learning framework for road marking extraction, classification and completion from mobile laser scanning point clouds. ISPRS Journal of Photogrammetry and Remote Sensing. 147, 178-192(2019).

【18】Yang M M, Wan Y C, Liu X L et al. Laser data based automatic recognition and maintenance of road markings from MLS system. Optics & Laser Technology. 107, 192-203(2018).

【19】Chen L C, Zhu Y K, Papandreou G et al. Encoder-decoder with atrous separable convolution for semantic image segmentation. ∥Ferrari V, Hebert M, Sminchisescu C, et al. Computer vision-ECCV 2018. Lecture notes in computer science. Cham: Springer. 11211, 833-851(2018).

【20】Chen L C, Papandreou G, Kokkinos I et al. DeepLab: semantic image segmentation with deep convolutional nets, atrous convolution, and fully connected CRFs. IEEE Transactions on Pattern Analysis and Machine Intelligence. 40(4), 834-848(2018).

【21】Gao L, Xie J and Cao T Z. An improved K-means clustering algorithm with high efficiency based on theKd tree. Computing Technology and Automation. 34(4), 69-74(2015).
高亮, 谢健, 曹天泽. 基于Kd树改进的高效K-means聚类算法. 计算技术与自动化. 34(4), 69-74(2015).

【22】Garcia-Garcia A, Orts-Escolano S, Oprea S et al. -04-22)[2017-04-22]. https:∥arxiv. org/abs/1704, (2017).

【23】Cheng M, Zhang H C, Wang C et al. Extraction and classification of road markings using mobile laser scanning point clouds. IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing. 10(3), 1182-1196(2017).

引用该论文

Gang Huang, Xianlin Liu. Automatic Extraction and Classification of Road Markings Based on Deep Learning[J]. Chinese Journal of Lasers, 2019, 46(8): 0804002

黄刚, 刘先林. 基于深度学习的道路标线自动提取与分类方法[J]. 中国激光, 2019, 46(8): 0804002

您的浏览器不支持PDF插件,请使用最新的(Chrome/Fire Fox等)浏览器.或者您还可以点击此处下载该论文PDF