基于多损失融合和混洗注意力的车载LiDAR点云道路标线提取方法

doi:10.13474/j.cnki.11-2246.2024.0823

测绘通报 ›› 2024, Vol. 0 ›› Issue (8): 135-140.doi: 10.13474/j.cnki.11-2246.2024.0823

• 技术交流 • 上一篇

基于多损失融合和混洗注意力的车载LiDAR点云道路标线提取方法

何银鑫^1,2, 齐华^2,3, 朱运权¹, 卢自来¹, 彭世勇², 刘洋^1,2

1. 四川省交通勘察设计研究院有限公司, 四川成都 610017;
2. 西南交通大学地球科学与环境工程学院, 四川成都 611756;
3. 高速铁路运营安全空间信息技术国家地方联合实验室, 四川成都 611756

收稿日期:2024-01-17 发布日期:2024-09-03
通讯作者: 刘洋。E-mail:ly895488739@my.swjtu.edu.cn
作者简介:何银鑫(1997—),男,硕士,助理工程师,主要研究方向为摄影测量与遥感。E-mail:heyinxin@my.swjtu.edu.cn
基金资助:
国家重点研发计划(2017YFB0503500);四川省科技厅重点研发项目(2021YFS0334);四川省交通勘察设计研究院有限公司科技项目(232022016)

Road marking extraction method from mobile LiDAR point clouds based on multi-loss fusion and shuffle attention

HE Yinxin^1,2, QI Hua^2,3, ZHU Yunquan¹, LU Zilai¹, PENG Shiyong², LIU Yang^1,2

1. Sichuan Communication Surveying and Design Institute Co., Ltd., Chengdu 610017, China;
2. Faculty of Geosciences and Environmental Engineering, Southwest Jiaotong University, Chengdu 611756, China;
3. State-Province Joint Engineering Laboratory in Spatial Information Technology for High-Speed Railway Safety, Chengdu 611756, China

Received:2024-01-17 Published:2024-09-03

摘要/Abstract

摘要： 道路标线的准确提取在高级辅助驾驶系统和高精度地图的开发中具有重要意义。针对现有的基于阈值的车载激光点云道路标线提取方法在反射强度与点密度分布不均、道路标线与路面对比度低时提取效果较差的问题,本文提出了基于多损失融合和混洗注意力的车载LiDAR点云道路标线提取方法。选取典型高速公路试验样区进行道路标线提取试验,并与常规方法进行了精度对比分析。试验表明,本文方法在道路标线提取精度方面优于其他方法,有望更好地服务于自动驾驶的高精度地图开发应用。

关键词: 多损失融合, 混洗注意力, 车载LiDAR点云, 道路标线提取

Abstract: The accurate extraction of road markings is of great significance in the development of advanced driving assistance system and high precision map. Since the point clouds have uneven distribution on reflection intensity and density or low contrast between road line and its surrounding road surface, the existing thresholding method is difficult to extract road line accurately,So this paper proposes the vehicle-mounted LiDAR point cloud road marking extraction method based on multi-loss fusion and mixing and shuffling attention, and selects a typical highway test sample area to conduct the road marking extraction test and compare and analyze the accuracy of the method with that of the conventional method. The accuracy comparison analysis is carried out with the conventional method. The test shows that the method in this paper is better than other methods in improving road marking extraction accuracy, which is expected to better serve the development and application of high-precision maps for autonomous driving.

Key words: multi-loss fusion, shuffle attention, mobile LiDAR point clouds, road marking extraction

中图分类号:

P237

何银鑫, 齐华, 朱运权, 卢自来, 彭世勇, 刘洋. 基于多损失融合和混洗注意力的车载LiDAR点云道路标线提取方法[J]. 测绘通报, 2024, 0(8): 135-140.

HE Yinxin, QI Hua, ZHU Yunquan, LU Zilai, PENG Shiyong, LIU Yang. Road marking extraction method from mobile LiDAR point clouds based on multi-loss fusion and shuffle attention[J]. Bulletin of Surveying and Mapping, 2024, 0(8): 135-140.

参考文献

[1] 姚连璧,秦长才,张邵华,等.车载激光点云的道路标线提取及语义关联[J].测绘学报,2020,49(4):480-488.
[2] SOILÁN M,SÁNCHEZ-RODRÍGUEZ A,DEL RÍO-BARRAL P,et al.Review of laser scanning technologies and their applications for road and railway infrastructure monitoring[J].Infrastructures,2019,4(4):58.
[3] WEN C,SUN X,LI J,et al.A deep learning framework for road marking extraction,classification and completion from mobile laser scanning point clouds[J].ISPRS Journal of Photogrammetry and Remote Sensing,2019,147:178-192.
[4] MA L,LI Y,LI J,et al.Generation of horizontally curved driving lines in HD maps using mobile laser scanning point clouds[J].IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing,2019,12(5):1572-1586.
[5] YAO L,QIN C,CHEN Q,et al.Automatic extraction and recognition of road markings based on vehicle laser point cloud[J].ISPRS Annals of Photogrammetry,Remote Sensing and Spatial Information Sciences,2020,5(2):313-325.
[6] 闫利,李赞.车载激光点云道路标线提取方法[J].遥感信息,2018,33(1):1-6.
[7] 李维刚,樊响,梅洋,等.基于车载激光点云的道路标线提取算法[J].中国激光,2022,49(7):0710001.
[8] GU J,WANG Z,KUEN J,et al.Recent advances in convolutional neural networks[J].Pattern Recognition,2018,77:354-377.
[9] 毛寅睿,陈雨人,刘洋东.基于车载激光点云的道路标线提取方法研究[J].公路交通科技,2019,36(7):127-135.
[10] 刘经南,詹骄,郭迟,等.智能高精地图数据逻辑结构与关键技术[J].测绘学报,2019,48(8):939-953.
[11] 刘经南,吴杭彬,郭迟,等.高精度道路导航地图的进展与思考[J].中国工程科学,2018,20(2):99-105.
[12] YANG B,FANG L,LI J.Semi-automated extraction and delineation of 3D roads of street scene from mobile laser scanning point clouds[J].ISPRS Journal of Photogrammetry and Remote Sensing,2013,79:80-93.
[13] KUMAR A,UPADHYAY N,GHOSAL P,et al.CSNet:a new DeepNet framework for ischemic stroke lesion segmentation[J].Comput Methods Programs Biomed,2020,193:105524.
[14] RONNEBERGER O,FISCHER P,BROX T.U-Net:convolutional networks for biomedicalimage segmentation[C]//Proceedings of 2015 International Conference on Medical Image Computing and Computer-assisted Intervention.[S.l.]:Springer,Cham,2015:234-241.
[15] ZHANG Q,YANG Y.SA-Net:shuffle attention for deep convolutional neural networks[C]//Proceedings of 2021 IEEE International Conference on Acoustics,Speech and Signal Processing.Toronto:IEEE,2021.
[16] LI X,HU X,YANG J.Spatial group-wise enhance:improving semantic feature learning in convolutional networks[J/OL].arXiv.2019-06-01.https://github.com/implus/PytorchInsight.
[17] 季顺平,魏世清.遥感影像建筑物提取的卷积神经元网络与开源数据集方法[J].测绘学报,2019,48(4):448-59.
[18] KUMAR P,MCELHINNEY C P,LEWIS P,et al.Automated road markings extraction from mobile laser scanning data[J].International Journal of Applied Earth Observation and Geoinformation,2014,32:125-137.

基于多损失融合和混洗注意力的车载LiDAR点云道路标线提取方法

Road marking extraction method from mobile LiDAR point clouds based on multi-loss fusion and shuffle attention

PDF

可视化

摘要/Abstract

引用本文

使用本文

参考文献

相关文章 4

编辑推荐

Metrics

本文评价

[1]	毛镜懿, 王竞雪, 董啸. 基于径向梯度的车载LiDAR点云路面点提取[J]. 测绘通报, 2024, 0(7): 48-54.
[2]	刘晋, 杨容浩, 文文, 谭骏祥, 兰青龙, 高祥, 汤洪. 一种车载LiDAR点云道路提取深度神经网络模型[J]. 测绘通报, 2023, 0(12): 8-12,18.
[3]	王艳军, 林云浩, 王书涵, 李少春, 王孟杰. OSM辅助车载LiDAR点云三维道路边界精细提取[J]. 测绘通报, 2022, 0(7): 18-25.
[4]	谢宏全, 汪秋玲, 蔡东健, 邢万里, 孙权, 刘付程. 利用车载LiDAR点云数据提取城市道路边界[J]. 测绘通报, 2021, 0(2): 64-67.