Point cloud object recognition method via histograms of dual deviation angle feature

被引:4
作者
Shi, Chunhao [1 ]
Wang, Chunyang [1 ,2 ]
Liu, Xuelian [2 ]
Sun, Shaoyu [1 ]
Xi, Guan [2 ]
Ding, Yueyang [1 ]
机构
[1] Changchun Univ Sci & Technol, Sch Elect & Informat Engn, Changchun, Peoples R China
[2] Xian Technol Univ, Xian Key Lab Act Photoelect Imaging Detect Technol, Xian, Peoples R China
基金
中国博士后科学基金;
关键词
HDDAF; point cloud; object recognition; local shape description; 3D; INFORMATION; DESCRIPTOR; EFFICIENT;
D O I
10.1080/01431161.2023.2214276
中图分类号
TP7 [遥感技术];
学科分类号
081102 ; 0816 ; 081602 ; 083002 ; 1404 ;
摘要
LiDAR point cloud object recognition is widely adopted in 3D perception applications and plays an important role in remote sensing, autonomous driving, robotics and other fields. We propose a novel point cloud object recognition method based on histograms of dual deviation angle feature, which we call the HDDAF. First, we use the normal alignment radial feature to sample feature points. Second, we construct the local reference frame (LRF) for the point cloud. Third, we calculate the connecting line between the barycentre and LRF origin, and calculate the deviation angle between connecting line and the LRF and deviation angle between connecting line and point cloud normal vectors. Fourth, we construct feature histograms of the two kinds of deviation angle. Finally, we construct a model library using the HDDAF algorithm to transform the point cloud data into a feature histogram, and we transform the point cloud of unknown objects into a feature histogram in the same way. We match the histogram of unknown objects with the model library histogram to recognize the unknown object. We rigorously test the proposed method on B3R and ModelNet40 standard datasets and real scenes, and compare with those of 10 other methods. For standard, noise and real scene data, the precisionrecall curve of the proposed method is superior to that of the comparison methods. The average computation times of the proposed method on the two datasets and real scenes are 0.221, 0.051 and 0.448 seconds, which are shorter than those of the comparison methods. The results show that the proposed method is highly descriptive, robust and computationally efficient.
引用
收藏
页码:3031 / 3058
页数:28
相关论文
共 50 条
[41]   A 3D Point Cloud Feature Identification Method Based on Improved Point Feature Histogram Descriptor [J].
Wang, Chunxiao ;
Xiong, Xiaoqing ;
Zhang, Xiaoying ;
Liu, Lu ;
Tan, Wu ;
Liu, Xiaojuan ;
Yang, Houqun .
ELECTRONICS, 2023, 12 (17)
[42]   A method of 3D object recognition and localization in a cloud of points [J].
Bielicki, Jerzy ;
Sitnik, Robert .
EURASIP JOURNAL ON ADVANCES IN SIGNAL PROCESSING, 2013,
[43]   A method of 3D object recognition and localization in a cloud of points [J].
Jerzy Bielicki ;
Robert Sitnik .
EURASIP Journal on Advances in Signal Processing, 2013
[44]   3D Object Detection Method with Image Semantic Feature Guidance and Cross-Modal Fusion of Point Cloud [J].
Li, Hui ;
Wang, Junyin ;
Cheng, Yuanzhi ;
Liu, Jian ;
Zhao, Guowei ;
Chen, Shuangmin .
Jisuanji Fuzhu Sheji Yu Tuxingxue Xuebao/Journal of Computer-Aided Design and Computer Graphics, 2024, 36 (05) :734-749
[45]   Point cloud coarse registration method based on local feature description [J].
Zhao, Chengli ;
Zhao, Zhangyan .
PHYSICA SCRIPTA, 2025, 100 (07)
[46]   Surface warpage detection method based on point cloud feature comparison [J].
Chen Y.-B. ;
Yi G.-D. ;
Zhang S.-Y. .
Zhejiang Daxue Xuebao (Gongxue Ban)/Journal of Zhejiang University (Engineering Science), 2021, 55 (01) :81-88
[47]   Semantic segmentation of large-scale point cloud scenes via dual neighborhood feature and global spatial-aware [J].
Liu, Tao ;
Ma, Tianen ;
Du, Ping ;
Li, Dehui .
INTERNATIONAL JOURNAL OF APPLIED EARTH OBSERVATION AND GEOINFORMATION, 2024, 129
[48]   Novel methods for noisy 3D point cloud based object recognition [J].
Xian-Feng Han ;
Xin-Yu Yan ;
Shi-Jie Sun .
Multimedia Tools and Applications, 2021, 80 :26121-26143
[49]   Point Cloud Object Recognition using 3D Convolutional Neural Networks [J].
Soares, Marcelo Borghetti ;
Wermter, Stefan .
2018 INTERNATIONAL JOINT CONFERENCE ON NEURAL NETWORKS (IJCNN), 2018,
[50]   A Novel Rock-Mass Point Cloud Registration Method Based on Feature Line Extraction and Feature Point Matching [J].
Liu, Lupeng ;
Xiao, Jun ;
Wang, Yunbiao ;
Lu, Zhengda ;
Wang, Ying .
IEEE TRANSACTIONS ON GEOSCIENCE AND REMOTE SENSING, 2022, 60