Gradient Enhancement Techniques and Motion Consistency Constraints for Moving Object Segmentation in 3D LiDAR Point Clouds

被引:0
|
作者
Tang, Fangzhou [1 ]
Zhu, Bocheng [1 ]
Sun, Junren [1 ,2 ]
机构
[1] Peking Univ, Sch Elect, Beijing 100871, Peoples R China
[2] China Univ Min & Technol Beijing, Sch Artificial Intelligence, Beijing 100083, Peoples R China
关键词
LiDAR point cloud; moving object segmentation; range image; gradient enhancement; motion consistency;
D O I
10.3390/rs17020195
中图分类号
X [环境科学、安全科学];
学科分类号
08 ; 0830 ;
摘要
The ability to segment moving objects from three-dimensional (3D) LiDAR scans is critical to advancing autonomous driving technology, facilitating core tasks like localization, collision avoidance, and path planning. In this paper, we introduce a novel deep neural network designed to enhance the performance of 3D LiDAR point cloud moving object segmentation (MOS) through the integration of image gradient information and the principle of motion consistency. Our method processes sequential range images, employing depth pixel difference convolution (DPDC) to improve the efficacy of dilated convolutions, thus boosting spatial information extraction from range images. Additionally, we incorporate Bayesian filtering to impose posterior constraints on predictions, enhancing the accuracy of motion segmentation. To handle the issue of uneven object scales in range images, we develop a novel edge-aware loss function and use a progressive training strategy to further boost performance. Our method is validated on the SemanticKITTI-based LiDAR MOS benchmark, where it significantly outperforms current state-of-the-art (SOTA) methods, all while working directly on two-dimensional (2D) range images without requiring mapping.
引用
收藏
页数:25
相关论文
共 50 条
  • [31] Drosophila-inspired 3D moving object detection based on point clouds
    Wang, Li
    Zhao, Dawei
    Wu, Tao
    Fu, Hao
    Wang, Zhiyu
    Xiao, Liang
    Xu, Xin
    Dai, Bin
    INFORMATION SCIENCES, 2020, 534 : 154 - 171
  • [32] OGC: Unsupervised 3D Object Segmentation from Rigid Dynamics of Point Clouds
    Song, Ziyang
    Yang, Bo
    ADVANCES IN NEURAL INFORMATION PROCESSING SYSTEMS 35 (NEURIPS 2022), 2022,
  • [33] 3D object recognition method with multiple feature extraction from LiDAR point clouds
    Tian, Yifei
    Song, Wei
    Sun, Su
    Fong, Simon
    Zou, Shuanghui
    JOURNAL OF SUPERCOMPUTING, 2019, 75 (08): : 4430 - 4442
  • [34] Investigation on Roof Segmentation for 3D Building Reconstruction from Aerial LIDAR Point Clouds
    Albano, Raffaele
    APPLIED SCIENCES-BASEL, 2019, 9 (21):
  • [35] 3D object recognition method with multiple feature extraction from LiDAR point clouds
    Yifei Tian
    Wei Song
    Su Sun
    Simon Fong
    Shuanghui Zou
    The Journal of Supercomputing, 2019, 75 : 4430 - 4442
  • [36] Fog Simulation on Real LiDAR Point Clouds for 3D Object Detection in Adverse Weather
    Hahner, Martin
    Sakaridis, Christos
    Dai, Dengxin
    Van Gool, Luc
    2021 IEEE/CVF INTERNATIONAL CONFERENCE ON COMPUTER VISION (ICCV 2021), 2021, : 15263 - 15272
  • [37] Urban 3D segmentation and modelling from street view images and LiDAR point clouds
    Babahajiani, Pouria
    Fan, Lixin
    Kamarainen, Joni-Kristian
    Gabbouj, Moncef
    MACHINE VISION AND APPLICATIONS, 2017, 28 (07) : 679 - 694
  • [38] RangeSeg: Range-Aware Real Time Segmentation of 3D LiDAR Point Clouds
    Chen, Tzu-Hsuan
    Chang, Tian Sheuan
    IEEE TRANSACTIONS ON INTELLIGENT VEHICLES, 2022, 7 (01): : 93 - 101
  • [39] Static map generation from 3D LiDAR point clouds exploiting ground segmentation?
    Arora, Mehul
    Wiesmann, Louis
    Chen, Xieyuanli
    Stachniss, Cyrill
    ROBOTICS AND AUTONOMOUS SYSTEMS, 2023, 159
  • [40] PointMatch: A consistency training framework for weakly supervised semantic segmentation of 3D point clouds
    Wu, Yushuang
    Yan, Zizheng
    Cai, Shengcai
    Li, Guanbin
    Han, Xiaoguang
    Cui, Shuguang
    COMPUTERS & GRAPHICS-UK, 2023, 116 : 427 - 436