越野环境下高精地图关键技术和应用展望

doi:10.12382/bgxb.2023.0854

摘要/Abstract

摘要：

随着人工智能等技术的发展,无人驾驶技术应运而生,越来越多的无人化装备也投入到作战应用中。面对复杂的越野环境,高精地图可以为无人车辆提供丰富的先验信息,辅助无人车辆进行环境感知、路径规划以及决策等,提升无人车辆越野机动能力。分析高精地图标准化、构建、应用等方面的研究现状,面向无人车辆越野环境下的自主机动任务需求,提出越野环境下高精地图的研究目标与技术体系,归纳总结越野环境下高精地图的基础理论与关键技术,并对越野环境下高精地图的应用发展进行了展望,为高精地图在无人驾驶方面的应用提供了参考。

关键词: 越野环境, 高精地图, 无人车辆, 自主导航

Abstract:

With the development of technologies such as artificial intelligence, the unmanned driving technology has emerged as the times require, and more and more unmanned equipment has also been put into combat applications. In the complex off-road environments, a high-definition map can provide rich prior information for unmanned ground vehicles, assist the unmanned ground vehicles in environmental perception, path planning and decision-making, and improve the off-road mobility of unmanned ground vehicles. This paper analyzes the research status of the standardization, construction and application of high-precision map, puts forward the research objectives and technical systems of high-precision map in off-road environments for the autonomous maneuvering tasks of unmanned vehicles in off-road environments, summarizes the basic theories and key technologies of high-precision map in off-road environments, and looks forward to the application and development of high-precision map in off-road environments, which provides a reference for the convenient application of high-precision map in unmanned driving.

Key words: off-road environments, high-definition map, unmanned ground vehicle, autonomous navigation

中图分类号:

TJ301

李兆冬, 赵熙俊, 杨婷婷, 齐晓隆, 周昌仪, 张立明. 越野环境下高精地图关键技术和应用展望[J]. 兵工学报, 2023, 44(S2): 1-11.

LI Zhaodong, ZHAO Xijun, YANG Tingting, QI Xiaolong, ZHOU Changyi, ZHANG Liming. Key Technologies and Application Prospects for High-definition Map in Off-road Environments[J]. Acta Armamentarii, 2023, 44(S2): 1-11.

图/表 11

图1 自主导航系统分层架构

Fig.1 Layered architecture of autonomous navigation system

表1 高精地图与传统电子地图主要区别

Table 1 The difference between high-definition map and traditional electronic map

地图种类	服务对象	精度	要素和属性
高精地图	无人车辆	亚米级	详细的车道模型(曲率、坡度、航向、限高、限重、限宽)、道路部件、定位地物等
传统电子地图	人	米级	道路、兴趣点

图2 OpenDRIVE道路结构和交叉口表示

Fig.2 Road structure and intersection representation of OpenDRIVE

表2 数据采集方法优缺点

Table 2 Advantages and disadvantages of data collection methods

数据来源	优势	劣势
众包采集	高效率、高实时性、高适应性	成本高、数据后处理代价较大
开源数据集	数据质量高、无需后处理	数据实时性差、针对性较差
MMS	数据可定制、成本较低、可针对特殊场景	地图覆盖范围小、数据实时性较差、数据采集耗时较多

图3 3D点云地图示意图

Fig.3 Schematic diagram of 3D point cloud map

图4 基于迭代增长生成路网(绿色表示路网结构)[37]

Fig.4 Generating road networks based on iterative growth(Green represents the road network structure)[37]

图5 “天-空-地”数据融合

Fig.5 Data fusion of space-air-ground

图6 地图要素提取分类效果

Fig.6 Effect of map feature extraction and classification

图7 实时点云框架

Fig.7 Framework for real-time point clouds

图8 无人机3D真彩色点云效果

Fig.8 Effect of UAV 3D true color point cloud

图9 无人车自主导航系统

Fig.9 Autonomous navigation system for unmanned vehicles

参考文献 61

[1]	王涛, 陈艳丽, 贾双成. 简述高精地图的特点[J]. 软件, 2018, 39(9):183-187.
	WANG T, CHEN Y L, JIA S C. Brief description of the characteristics of high precision map[J]. Computer Engineering & Software, 2018, 39(9):183-187. (in Chinese)
[2]	王彬. 基于多传感器融合的高精度地图生成方法研究[D]. 重庆: 重庆大学, 2022:1-6.
	WANG B. Research on high-precision map generation method based on multi-sensor fusion[D]. Chongqing: Chongqing University, 2022:1-6. (in Chinese)
[3]	CHEN L, LI Y C, CHAO H, et al. Milestones in autonomous driving and intelligent vehicles: survey of surveys[J]. IEEE Transactions on Intelligent Vehicles, 2023, 8(2):1046-1056. doi: 10.1109/TIV.2022.3223131 URL
[4]	SHIM I, CHOI J, SHIN S, et al. An autonomous driving system for unknown environments using a unified map[J]. IEEE Transactions on Intelligent Transportation Systems, 2015, 16(4): 1999-2013. doi: 10.1109/TITS.2015.2389237 URL
[5]	JO K, SUNWOO M. Generation of a precise roadway map for autonomous cars[J]. IEEE Transactions on Intelligent Transportation Systems, 2014, 15(3):925-937. doi: 10.1109/TITS.2013.2291395 URL
[6]	DIAZ-DIAZ A, OCAÑA M, LLAMAZARES Á, et al. HD maps: exploiting OpenDRIVE potential for path planning and map monitoring[C]// Proceedings of the 2022 IEEE Intelligent Vehicles Symposium (IV). Aachen, Germany: IEEE, 2022:1211-1217.
[7]	LIN J R, LIU X Y, ZHANG F. A decentralized framework for simultaneous calibration, localization and mapping with multiple LiDARs[C]// Proceedings of the 2020 IEEE RSJ International Conference on Intelligent Robots and Systems. Las Vegas, NV, US: IEEE, 2020:4870-4877.
[8]	MI L, ZHAO H, NASH C, et al. HDMapGen: A hierarchical graph generative model of high definition maps[C]// Proceedings of the 2022 IEEE Conference on Computer Vision and Pattern Recognition. Nashville, TN, US: IEEE, 2021:4225-4234.
[9]	王淞. 特定场景轻量级高精地图设计[D]. 重庆: 重庆邮电大学, 2022.
	WANG S. Lightweight high-precision map for specific scenes[D]. Chongqing: Chongqing University of Posts and Telecommunications, 2022. (in Chinese)
[10]	吕钊凤, 李峥嵘. 无人驾驶,地图先行——盘点中国19家高精度地图厂商[J]. 智能网联汽车, 2019(4):22-28.
	LÜ Z F, LI Z R. Unmanned driving, map first-inventory of 19 high definition map manufacturers in China[J]. Intelligent Connected Vehicles, 2019(4):22-28. (in Chinese)
[11]	黄琛, 尹彤, 王建明. 高精度地图标准化建设探讨[J]. 中国标准化, 2021(21):89-93.
	HUANG C, YIN T, WANG J M. Discussion on the standardization of high definition map[J]. China Standardization, 2021(21):89-93. (in Chinese)
[12]	POGGENHANS F, PAULS J, JANOSOVITS J, et al. Lane-let2: a high-definition map framework for the future of automated driving[C]// Proceedings of the 2018 21st International Conference on Intelligent Transportation Systems. Maui, HI, US: IEEE, 2018:1672-1679.
[13]	ALTHOFF M, URBAN S, KOSCHI M. Automatic conversion of road networks from OpenDRIVE to lanelets[C]// Proceedings of the 2018 IEEE International Conference on Service Operations and Logistics, and Informatics (SOLI). Singapore: IEEE, 2018:157-162.
[14]	DUPUIS M, STROBL M, GREZLIKOWSKI H. OpenDRIVE 2010 and beyond-status and future of the de facto standard for the description of road networks[C]// Proceedings of 2010 Driving Simulation Conference Europe. Paris, France: IEEE, 2010:231-242.
[15]	ISO/TC204. Intelligent transport systems—geographic data files(GDF)—GDF 5.1-part 2: map data used in automated driving systems, cooperative ITS, and multimodal transport: ISO 20524-2:2020[S]. Warrendale, PA, US: International Standards Organization, 2020.
[16]	ISO/TC204. Intelligent transport systems—geographic data files(GDF)—GDF5.0: ISO 14825:2011[S]. Warrendale, PA, US: International Standards Organization, 2011.
[17]	JUNKER M. Announcement: 1st NDS open lane day on May 11, 2017[EB/OL]. (2017-04-11)[2020-07-17]. http://www.openlanemodel.org/news_openlaneday.php.
[18]	周勇, 刘尚魁. 构建基于 Appollo 的高精度地图解决方案[J]. 电子技术与软件工程, 2018(21):139.
	ZHOU Y, LIU S K. Building a high definition map solution based on Apollo[J]. Electronic Technology & Software Engineering, 2018(21):139. (in Chinese)
[19]	张晓华. 高精度地图为自主泊车保驾护航[C]// 卫星导航定位与北斗系统应用 2019——北斗服务全球融合创新应用. 北京: 测绘出版社, 2019:130-133.
	ZHANG X H. High Definition map provides protection for autonomous parking[C]// Proceedings of 2019 Satellite Navigation Positioning and Beidou System Application 2019-Global Integration and Innovative Application of Beidou Service. Beijing: Mapping Press, 2019:130-133. (in Chinese)
[20]	HOUSTON J, ZUIDHOF G, BERGAMINI L, et al. One thousand and one hours: self-driving motion prediction dataset[C]// Proceedings of the 4th Conference on Robot Learning.Virtual, Online, US: ML Research Press, 2020:409-418.
[21]	MARCHANT A. Behind the map:how we keep our maps up to date[EB/OL].(2019-10-22)[2023-03-15]. https://www.tomtom.com/newsroom/behind-the-map/continuous-map-processing/.
[22]	HERE, 2021. Map data \| static map API[Online]. Available: https://www.here.com/platform/map-data.
[23]	GEIGER A, LENZ P, STILLER C, et al. Vision meets robotics: the KITTI dataset[J]. The International Journal of Robotics Research, 2013, 32(11): 1231-1237. doi: 10.1177/0278364913491297 URL
[24]	AESAR H, BANKITI V, LANG A, et al. nuScenes: a multimodal dataset for autonomous driving[C]// Proceedings of the 2020 IEEE Conference on Computer Vision and Pattern Recognition (CVPR). Seattle, WA, US: IEEE, 2020:11618-11628.
[25]	ZHANG J, SINGH S. Loam: lidar odometry and mapping in real-time[C]// Proceedings of the 2014 IEEE Robotics:Science and Systems. Berkeley, CA, US: MIT, 2014:1-7.
[26]	SHAN T, ENDLOT B. LeGO-LOAM: lightweight and ground-optimized LiDAR odometry and mapping on variable terrain[C]// Proceedings of the 2018 IEEEInternational Conference on Intelligent Robots and Systems(IROS). Madrid, Spain:IEEE, 2018:4758-4765.
[27]	SHAN T, ENGLOT B, MEYERS D, et al. LIO-SAM: tightly-coupled LiDAR inertial odometry via smoothing and mapping[C]// Proceedings of the 2020 IEEE International Conference on Intelligent Robots and Systems (IROS).Las Vegas, NV, US: IEEE, 2020:5135-5142.
[28]	XU W, ZHANG F. FAST-LIO: a fast, robust LiDAR-inertial odometry package by tightly-coupled iterated Kalman filter[J]. IEEE Robotics and Automation Letters, 2021, 6(2): 3317-3324. doi: 10.1109/LRA.2021.3064227 URL
[29]	XU W, CAI Y X, HE D J, et al. FAST-LIO2: fast direct LiDAR-inertial odometry[J]. IEEE Transactions on Robotics, 2022, 38(4):2053-2073. doi: 10.1109/TRO.2022.3141876 URL
[30]	SHAN T, ENGLOT B, RATTI C, et al. LVI-SAM: tightly-coupled LiDAR-visual-tnertial odometry via smoothing and mapping[C]// Proceedings of the 2021 IEEE International Conference on Robotics and Automation. Xi’an, China: IEEE, 2021:5692-5698.
[31]	GHANDRTH H, BOULILA W, MASOOD S, et al. Semantic segmentation and edge detection—approach to road detection in very high resolution satellite images[J]. Remote Sens, 2022, 14(3):613. doi: 10.3390/rs14030613 URL
[32]	CHEN L, PAPANDREOU G, SCHROFF F, et al. Rethinking atrous convolution for semantic image segmentation[C]// Proceedings of the 2020 IEEE International Symposium on Electronic Imaging Science and Technology. Burlingame, CA, US: IEEE, 2020:1-11.
[33]	CHIKHAOUI B, RUER P, ÉVELYNE F, et al. A deep learning method for automatic visual attention detection in older drivers[C]// Proceedings of the 17th International Conference on Smart Living and Public Health. New York, NY, US: Springer Science and Business Media Deutschland GmbH, 2019:49-60.
[34]	LI X, ZHANG W, DING Q. Understanding and improving deep learning-based rolling bearing fault diagnosis with attention mechanism Signal Process[J]. Signal Processing, 2019, 161:136-154. doi: 10.1016/j.sigpro.2019.03.019 URL
[35]	ZANG Z Y, YANG X H, FENG Z Q. Self-attention mechanism fusion for recommendation in heterogeneous information network[C]// Proceedings of the 8th International Conference on Signal and Information Processing, Network and Computers. Jinan, China: Springer Science and Business Media Deutschland GmbH, 2022, 917:1147-1155.
[36]	WANG J, WEI X, LIU X. LMSA: Low-relation multi-head self-attention mechanism in visual transformer[C]// Proceedings of the 2nd International Conference on Image, Vision and Intelligent Systems. Jinan, China: Springer Science and Business Media Deutschland GmbH, 2022, 1019:609-619.
[37]	BASTANI F, HE S T, ABBAR S, et al. RoadTracer: automatic extraction of road networks from aerial images[C]// Proceedings of the 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition. Salt Lake City, UT, US: IEEE Computer Society, 2018:4720-4728.
[38]	IBRAHIM M, AKHTAR N, JALWANA M, et al. High definition LiDAR mapping of perth CBD[C]// Proceedings of 2021 Digit Image Comput:Techniques and Applications. Gold Coast, Australia: IEEE, 2021:1-8.
[39]	DING S D, FU Y J, WANG W F, et al. Creation of high definition map for autonomous driving within specific scene[C]// Proceedings of the 2021 SPIE International Conference on Smart Transportation and City Engineering 2021. Chongqing, China:SPIE, 2021:1365-1373.
[40]	GU S, ZHANG YG, TANG J H, et al. Road detection through CRF based LiDAR-camera fusion[C]// Proceedings of the 2019 International Conference on Robotics and Automation. Montreal, QC, Canada: IEEE, 2019:3832-3838.
[41]	GU S, ZHANG Y G, YANG J H, et al. Two-view fusion based convolutional neural network for urban road detection[C]//Proceedings of the 2019 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS). Macau, China:IEEE, 2019:6144-6149.
[42]	PLACHETKA C, FRICKE J. KLINGNER M, et al. DNN-based recognition of pole-like objects in LiDAR point clouds[C]// Proceedings of the 2021 IEEE International Intelligent Transportation Systems Conference. Indianapolis, IN, US:IEEE, 2021:2889-2896.
[43]	仲跻炜, 陈大吾, 朱峰. 基于北斗高精度定位和车道级高精度地图的车辆监控平台设计与实现[J]. 全球定位系统, 2015, 40(4):50-55.
	ZHONG J W, CHEN D W, ZHU F. Design and implementation of a vehicle monitoring platform based on beidou high precision positioning and lane level high precision maps[J]. GNSS World of China, 2015, 40(4):50-55. (in Chinese)
[44]	李亮. 基于点云匹配的智能车定位方法研究[D]. 上海: 上海交通大学, 2018:1-10.
	LI L. Research on point cloud matching based localization for intelligent vehicles[D]. Shanghai: Shanghai Jiao Tong University, 2018:1-10. (in Chinese)
[45]	王一文, 钱闯, 唐健, 等. 预建高精度地图的封闭区域UGV 自动驾驶导航定位[J]. 测绘通报, 2020(1):21-25. doi: 10.13474/j.cnki.11-2246.2020.0005
	WANG Y W, QIAN C, TANG J, et al. UGV’s automated driving navigation and localization in closed areas with pre-established high precision planar map[J]. Bulletin of Surveying and Mapping, 2020(1):21-25. (in Chinese)
[46]	周涛. 单目视觉下车道线检测与跟踪算法的研究与测试[D]. 武汉: 华中师范大学, 2016:1-5.
	ZHOU T. Research and testing on lane detection and tracking algorithm based on monocular vision[D]. Wuhan: Central China Normal University, 2016:1-5. (in Chinese)
[47]	胡钊政, 孙莹妹, 李祎承. 路面路标高精度地图构建与多尺度车辆定位[J]. 哈尔滨工业大学学报, 2019, 51(9):149-156.
	HU Z Z, SUN Y S, LI Y C. High definition map construction from pavement landmarks for multi-scale vehicle localization[J]. Journal of Harbin Institute of Technology, 2019, 51(9): 149-156. (in Chinese)
[48]	祖似杰, 张攀, 罗跃军. 基于车道驾驶态势的拓扑构建与路径规划[J]. 地理空间信息, 2018, 16(5):53-56.
	ZU S J, ZHANG P, LUO Y J. Topology construction and path planning based on lane driving situation[J]. Geospatial Information, 2019, 16(5):53-56. (in Chinese)
[49]	訾烨, 任明武. 一种基于高精度地图匹配误差的路径规划方法[J]. 计算机与数字工程, 2021, 49(11):2248-2253.
	ZI Y, REN M W. A path planning method based on high precision map matching error[J]. Computer & Digital Engineering, 2021, 49(11):2248-2253. (in Chinese)
[50]	陈根. 移动测量技术在高精度地图数据生产方面的应用分析[J]. 城市建设理论研究(电子版), 2018(5):104-105.
	CHEN G. Application analysis of mobile measurement technology in high-precision map data production[J]. Theoretical Research in Urban Construction, 2018(5):104-105. (in Chinese)
[51]	曹波. 基于多传感器的高精度地图生成技术研究[J]. 内燃机与配件, 2019(16):133-135.
	CAO B. Research on high precision map generation technology based on multiple sensors[J]. Internal Combustion Engine & Parts, 2019(16):133-135. (in Chinese)
[52]	ZHANG J Y. Application of panoramic laser mobile measurement in highway operation management[C]// Proceedings of the 2021 IEEE 2021 4th International Symposium on Traffic Transportation and Civil Architecture. Suzhou, China:IEEE, 2021:7-10.
[53]	CHEN W F, ZHOU C J, SHANG G T, et al. SLAM Overview: from single sensor to heterogeneous fusion[J]. Remote Sensing, 2022, 14(23):6033. doi: 10.3390/rs14236033 URL
[54]	WU W W. Urban road automatic mapping method based on remote sensing image[C]// Proceedings of the 2023 IEEE 4th Annual Flagship India Council International Subsections Conference:Computational Intelligence and Learning Systems 2023. Mysore, India:IEEE, 2023:1-6.
[55]	LI X, LI H W. Application of neural network in map annotation recognition[C]// Proceedings of the 2nd International Conference on Artificial Intelligence, Big Data and Algorithms. Virtual, Online: VDE VERLAG GmbH, 2022:76-79.
[56]	XU H H, LI W, CAI Z P. Analysis on methods to effectively improve transfer learning performance[J]. Theoretical Computer Science, 2023, 940(B):90-107.
[57]	LIU Q, ZHANG Y R, WANG H X. EdgeMap: CrowdSourcing high definition map in automotive edge computing[C]//Proceedings of the 2022 IEEE International Conference on Communications. Seoul, Korea:IEEE, 2022:4300-4305.
[58]	PORTILLA-FIGUERAS A, LLOPIS-SANCHEZ S, JIMENEZ-FERNANDEZ S, et al. Examining 5G technology-based applications for military communications[C]// Proceedings of the 2023 IEEE Lecture Notes in Computer Science (including Subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics). Copenhagen, Denmark: Springer Science and Business Media Deutschland GmbH, 2023:449-465.
[59]	LI F D, ZENG Q H, XIONG Z, et al. Development and future direction of unmanned system based on inertial integrated navigation[C]// Proceedings of the 2023 IEEE Lecture Notes in Electrical Engineering. Harbin, China: Springer Science and Business Media Deutschland GmbH, 2023:7152-7161.
[60]	YANG B, SONG X W, GAO Z H. A lane level bi-directional hybrid path planning method based on high definition map[J]. World Electric Vehicle Journal, 2021, 12(4):227. doi: 10.3390/wevj12040227 URL
[61]	QU Y Q, FAN Y X, ZHANG X L, et al. Improved A^* path planning method based on the grid map[J]. Sensors, 2022, 22(16):6198. doi: 10.3390/s22166198 URL