KDD: A kernel density based descriptor for 3D point clouds

被引:19
|
作者
Zhang, Yuhe [1 ]
Li, Chunhui [1 ]
Guo, Bao [1 ]
Guo, Chenhao [1 ]
Zhang, Shunli [1 ]
机构
[1] Northwest Univ, Sch Informat Sci & Technol, Xian, Peoples R China
基金
中国国家自然科学基金;
关键词
3D feature descriptor; Kernel density estimation; Point cloud registration; KL divergence; OBJECT RECOGNITION; SURFACE; REPRESENTATION; REGISTRATION; EFFICIENT; IMAGES;
D O I
10.1016/j.patcog.2020.107691
中图分类号
TP18 [人工智能理论];
学科分类号
081104 ; 0812 ; 0835 ; 1405 ;
摘要
3D feature description is one of the central techniques that rely on point clouds since a lot of point cloud processing techniques apply the point-to-point correspondences that are achieved via feature descriptors as input data. The feature descriptor encodes the information of the underlying surface around the feature point so as to make a local surface distinguished from another. The focus of the existing descriptors is accumulating the geometric or topological measurements into histograms or encoding the 2D images that are acquired by rotationally projecting the 3D local surfaces onto 2D planes. Histograms can hardly deal with three or more dimensional information, and the rotational projection operation does bring much unnecessary intermediate computations. To overcome these limitations, in this article, a descriptor named Kernel Density Descriptor (KDD) has been presented. One core contribution of this method is to encode the information of the whole 3D space around the feature point via kernel density estimation, and another is providing the strategy for selecting different matching metrics for datasets with diverse levels of resolution qualities. We compare KDD against several representative descriptors on publicly available datasets, the experimental results demonstrate that the KDD descriptor achieves a satisfactory and balanced performance in terms of descriptiveness, robustness, and compactness, furthermore, the comparisons validate the overall superiority of our method. The benefits and applicability on object registration and recognition and 3D object reconstruction are demonstrated by the favorable results that are obtained for both public datastes and the real-world point clouds of Terracotta fragments. (C) 2020 Elsevier Ltd. All rights reserved.
引用
收藏
页数:13
相关论文
共 50 条
  • [1] Object Classification Based on 3D Point Clouds Covariance Descriptor
    Zhang, Heng
    Zhuang, Bin
    Liu, Yanli
    [J]. 2017 IEEE INTERNATIONAL CONFERENCE ON COMPUTATIONAL SCIENCE AND ENGINEERING (CSE) AND IEEE/IFIP INTERNATIONAL CONFERENCE ON EMBEDDED AND UBIQUITOUS COMPUTING (EUC), VOL 2, 2017, : 234 - 237
  • [2] ICP registration with DCA descriptor for 3D point clouds
    He, Ying
    Yang, Jun
    Hou, Xingming
    Pang, Shiyan
    Chen, Jia
    [J]. OPTICS EXPRESS, 2021, 29 (13) : 20423 - 20439
  • [3] Registration of 3D point clouds using a local descriptor based on grid point normal
    Wang, Jiang
    Wu, Bin
    Kang, Jiehu
    [J]. APPLIED OPTICS, 2021, 60 (28) : 8818 - 8828
  • [4] An Improved Local Descriptor based Object Recognition in Cluttered 3D Point Clouds
    Liu, X.
    Lu, Y.
    Wu, T.
    Yuan, T.
    [J]. INTERNATIONAL JOURNAL OF COMPUTERS COMMUNICATIONS & CONTROL, 2018, 13 (02) : 221 - 234
  • [5] ICP Registration Base on 3D Point Clouds Feature Descriptor
    He, Ying
    Yang, Jun
    Li, Zhiheng
    Liang, Bin
    [J]. TWELFTH INTERNATIONAL CONFERENCE ON GRAPHICS AND IMAGE PROCESSING (ICGIP 2020), 2021, 11720
  • [6] BRoPH: An efficient and compact binary descriptor for 3D point clouds
    Zou, Yu
    Wang, Xueqian
    Zhang, Tao
    Liang, Bin
    Song, Jingyan
    Liu, Houde
    [J]. PATTERN RECOGNITION, 2018, 76 : 522 - 536
  • [7] Discrete Geodesic Distribution-Based Graph Kernel for 3D Point Clouds
    Balci, Mehmet Ali
    Akguller, Omer
    Batrancea, Larissa M.
    Gaban, Lucian
    [J]. SENSORS, 2023, 23 (05)
  • [8] 3D Feature Detector-Descriptor Pair Evaluation on Point Clouds
    Stancelova, Paula
    Sikudova, Elena
    Cernekova, Zuzana
    [J]. 28TH EUROPEAN SIGNAL PROCESSING CONFERENCE (EUSIPCO 2020), 2021, : 590 - 594
  • [9] Spherical Kernel for Efficient Graph Convolution on 3D Point Clouds
    Lei, Huan
    Akhtar, Naveed
    Mian, Ajmal
    [J]. IEEE TRANSACTIONS ON PATTERN ANALYSIS AND MACHINE INTELLIGENCE, 2021, 43 (10) : 3664 - 3680
  • [10] An enhanced Density-based spatial separation strategy of 3D point clouds
    Kim, Jongwon
    Kim, Jinsoo
    Cho, Jeongho
    [J]. 2019 13TH INTERNATIONAL CONFERENCE ON SIGNAL PROCESSING AND COMMUNICATION SYSTEMS (ICSPCS), 2019,