A dynamic clustering algorithm based on the point clouds distribution characteristics of obstacle

doi:10.3969/j.issn.1674-8484.2024.02.015

Abstract

Abstract:

The lidar sensor plays an important role in the object detection of automatic driving systems, but the spatial distribution of point cloud is uneven because of its scanning mechanism, in which case a bunch of erroneous is yielded by the conventional clustering algorithms with fixed parameters. To solve such problems, a dynamic clustering algorithm based on the distribution characteristic of object point clouds was proposed, using the elliptical shape as the spatial neighborhood which adjusted its size according to the position of the sampling points. The key parameters were further designed quantitatively with the KITTI dataset considering comprehensive clustering performances, and the comparison experiment was conducted on campus. The results show that the proposed dynamic clustering algorithm can effectively reduce the erroneous results, such as 70.60% of over-clustering and 49.76% of under-clustering, caused by the fixed neighborhoods of density-based spatial clustering of applications with density-based spatial clustering of applications with noise(DBSCAN), therefore, effectively enhancing the comprehansive clustering performance of the algorithm.

Key words: autonomous driving, object detection, lidar sensor, point cloud clustering, KITTI dataset, density-based spatial clustering of applications with noise (DBSCAN)

CLC Number:

U471

LI Caihong, HE Chenyang, GAO Feng, CHEN Jiaxin. A dynamic clustering algorithm based on the point clouds distribution characteristics of obstacle[J]. Journal of Automotive Safety and Energy, 2024, 15(2): 261-267.

Figures/Tables 10

References 34

[1]	GAO Feng, HAN Yu, LI Shengbo, et al. Accurate pseudospectral optimization of nonlinear model predictive control for high-performance motion planning[J]. *IEEE Trans Intel Vehi*, 2023, 8(2): 1034-1045.
[2]	Jebamikyous H H, Kashef R. Autonomous vehicles perception (AVP) using deep learning: modeling, assessment, and challenges[J]. *IEEE Access*, 2022, 10: 10523-10535. doi: 10.1109/ACCESS.2022.3144407 URL
[3]	Alaba S Y, Ball J E. A Survey on deep-learning-based LiDAR 3D object detection for autonomous driving[J]. *Sensors*, 2022, 22(24): 9577. doi: 10.3390/s22249577 URL
[4]	LI You, Ibanez-guzman J. Lidar for autonomous driving: The principles, challenges, and trends for automotive lidar and perception systems[J]. *IEEE Sign Proc Maga*, 2020, 37(4): 50-61.
[5]	XIE Yuxing, TIAN Jiaojiao, ZHU Xiaoxiang. Linking points with labels in 3D: A review of point cloud semantic segmentation[J]. *IEEE Geosci Remo Sens Maga*, 2020, 8(4): 38-59.
[6]	SUN Pengpeng, SUN Chenghao, WAN Lingfeng, et al. Objects detection with 3-D roadside LIDAR under snowy weather[J]. *IEEE Sens J*, 2022, 22(23): 23051-23063. doi: 10.1109/JSEN.2022.3215768 URL
[7]	SU Hang, Maji S, Kalogerakis E, et al. Multi-view convolutional neural networks for 3D shape recognition[C]// 2015 IEEE Int’l Conf Comp Visi (ICCV). Santiago, Chile. 2015.
[8]	Lang A H, Vora S, Caesar H, et al. Pointpillars: Fast encoders for object detection from point clouds[C] // 2019 IEEE/CVF Conf Comp Visi Patt Recog (CVPR). California, USA. 2019.
[9]	DENG Jiajun, SHI Shaoshuai, LI Peiwei, et al. Voxel R-CNN: Towards high performance voxel-based 3D object detection[J]. *Proceed AAAI Conf Artif Intel*, 2021, 35(2): 1201-1209.
[10]	YAN Yan, MAO Yuxing, LI Bo. SECOND: Sparsely embedded convolutional detection[J]. *Sensors*, 2018, 18(10): 3337. doi: 10.3390/s18103337 URL
[11]	SHI Weijing, Ragunathan, Rajkumar. Point-GNN: Graph neural network for 3D object detection in a point cloud[C]// 2020 IEEE/CVF Conf Comp Visi Patt Recog (CVPR). Washington, USA. 2020.
[12]	YANG Honghui, LIU Zili, WU Xiaopei, et al. Graph R-CNN: Towards accurate 3D object detection with semantic-decorated local graph[C]// Comp Visi-ECCV 2022: 17th European Conf. Tel Aviv, Israel. 2022.
[13]	ZHENG Wu, TANG Weiliang, JIANG Li, et al. SE-SSD: Self-ensembling single-stage object detector from point cloud[C]// Proceed IEEE/CVF Conf Comp Visi Patt Recog. 2021: 14494-14503.
[14]	ZHANG Linfeng, DONG Runpei, TAI Hung-Shuo, et al. Pointdistiller: Structured knowledge distillation towards efficient and compact 3d detection[C]// Proceed IEEE/CVF Conf Comp Visi Patt Recog. Vancouver, BC, Canada, 2023.
[15]	Qi C R, SU Hao, MO Kaichun, et al. PointNet: Deep learning on point sets for 3D classification and segmentation[C]// 2017 IEEE Conf Comp Visi Patt Recog (CVPR), Hawaii, USA. 2017.
[16]	Rahman M M, TAN Yanhao, XUE Jian, et al. Notice of violation of IEEE publication principles: Recent advances in 3D object detection in the era of deep neural networks: A survey[J]. *IEEE Trans Imag Proce*, 2020, 29: 2947-2962.
[17]	GUO Ente, CHEN Zhifeng, Rahardja S, et al. 3D Detection and pose estimation of vehicle in cooperative vehicle infrastructure system[J]. *IEEE Sens J*, 2021, 21(19): 21759-21771. doi: 10.1109/JSEN.2021.3101497 URL
[18]	LI Xinggang, ZHANG Yaping, YANG Yuwei. Outlier detection for reconstructed point clouds based on image[C]// 2017 First Int’l Conf Elect Instrum Info Syst. Harbin, China. 2017.
[19]	ZHANG Zhengyu, JIN Mengdi. AOMC: An adaptive point cloud clustering approach for feature extraction[J]. *Scientific Programming*, 2022, 2022: e3744086.
[20]	SONG Yanjie, ZHANG Han, LIU Yuanqiang, et al. Background filtering and object detection with a stationary LiDAR using a layer-based method[J]. *IEEE Access*, 2020, 8: 184426-184436. doi: 10.1109/Access.6287639 URL
[21]	WU Jianqing, XU Hao, ZHENG Jianying, et al. Automatic vehicle detection with roadside LiDAR data under rainy and snowy conditions[J]. *IEEE Intel Transport Syst Maga*, 2021, 13(1): 197-209.
[22]	ZHAO Junxuan, XU Hao, LIU Hongchao, et al. Detection and tracking of pedestrians and vehicles using roadside LiDAR sensors[J]. **Transport Res Part C: Emerg Tech**, 2019, 100: 68-87. doi: 10.1016/j.trc.2019.01.007 URL
[23]	CHIANG Yenhung, HSU Chihming, Tsai A. Fast multi-resolution spatial clustering for 3D point cloud data[C]// 2019 IEEE Int’l Conf Syst, Man Cybe (SMC). Bari, Italy. 2019.
[24]	JIANG Wuhua, SONG Chuanzheng, WANG Hai, et al. Obstacle detection by autonomous vehicles: An adaptive neighborhood search radius clustering approach[J]. *Machines*, 2023, 11(1): 54. doi: 10.3390/machines11010054 URL
[25]	CHEN Jingrong, XU Hao, WU Jianqing, et al. Deer crossing road detection with roadside LiDAR sensor[J]. *IEEE Access*, 2019, 7: 65944-65954. doi: 10.1109/ACCESS.2019.2916718
[26]	CHEN Xiao, CHEN Zhuang, LIU Guoxiang, et al. Railway overhead contact system point cloud classification[J]. *Sensors*, 2021, 21(15): 4961. doi: 10.3390/s21154961 URL
[27]	Athanesious J J, Vasuhi S, Vaidehi V, et al. Adaptive density based data mining technique for detection of abnormalities in traffic video surveillance[J]. *J Intel Fuzz Syst*, 2020, 39(3): 3737-3747.
[28]	LIU Kaiqi, WANG Jianqiang. Fast dynamic vehicle detection in road scenarios based on pose estimation with convex-Hull model[J]. *Sensors*, 2019, 19(14): 3136. doi: 10.3390/s19143136 URL
[29]	MENG Qinghao, WANG Wenguan, ZHOU Tianfei, et al. Towards a weakly supervised framework for 3D point cloud object detection and annotation[J]. *IEEE Trans Patt Anal Mach Intel*, 2022, 44(8): 4454-4468.
[30]	Geiger A, Lenz P, Stiller C, et al. Vision meets robotics: The KITTI dataset[J]. *Int’l J Robot Res*, 2013, 32(11): 1231-1237.
[31]	JIN Xianjian, YANG Hang, LIAO Xin, et al. A robust gaussian process-based LiDAR ground segmentation algorithm for autonomous driving[J]. *Machines*, 2022, 10(7): 507. doi: 10.3390/machines10070507 URL
[32]	Laghmara H, Laurain T, Cudel C, et al. 2.5D evidential grids for dynamic object detection[C]// 2019 22th Int’l Conf Info Fusion (FUSION). Ontario, Canada. 2019.
[33]	SONG Weilong, XIONG Guangming, CHEN Huiyan. Intention-aware autonomous driving decision-making in an uncontrolled intersection[J]. *Math Prob Engi*, 2016, 2016: 1-15.
[34]	Intempora. RTMaps [/OL]. 2023. [2023-01-03]. https://intempora.com/products/rtmaps/.

划分区域	X轴距离/ m	MinPts	邻域半径/ m
区域1	[1, 20)	8	0.47
区域2	[18 40)	5	0.98
区域3	[38, 80]	4	1.83

划分区域	X轴距离/ m	MinPts	邻域半径/ m
区域1	[1, 20)	8	0.47
区域2	[18 40)	5	0.98
区域3	[38, 80]	4	1.83