Point2Wave: 3-D Point Cloud to Waveform Translation Using a Conditional Generative Adversarial Network With Dual Discriminators

被引:1
作者
Shinohara, Takayuki [1 ]
Xiu, Haoyi [2 ]
Matsuoka, Masashi [2 ]
机构
[1] Tokyo Inst Technol, Dept Architecture & Bldg Engn, Yokohama, Kanagawa 2268502, Japan
[2] Tokyo Inst Technol, Yokohama, Kanagawa 2268502, Japan
关键词
Three-dimensional displays; Laser radar; Atmospheric modeling; Superresolution; Deep learning; Generators; Feature extraction; Airborne LiDAR; conditional generative adversarial network; deep learning; full waveform LiDAR; LAND-COVER CLASSIFICATION; AIRBORNE LIDAR DATA; SUPERRESOLUTION;
D O I
10.1109/JSTARS.2021.3124610
中图分类号
TM [电工技术]; TN [电子技术、通信技术];
学科分类号
0808 ; 0809 ;
摘要
Since 2017, many deep learning methods for 3-D point clouds observed by airborne LiDAR (airborne 3-D point clouds) have been proposed. Moreover, not only a deep learning method for airborne 3-D point clouds but also a deep learning method for points and their waveforms observed by full-waveform LiDAR (airborne FW data) was proposed. We need to achieve highly accurate land cover classification by using airborne FW data, but open data often only have airborne 3-D point clouds available. Therefore, to improve the performance of land cover classification when using airborne 3-D point clouds published as open data, it is important to restore waveforms from airborne 3-D point clouds. In this article, we propose a deep learning model to translate an airborne 3-D point cloud to airborne FW data (called a point-to-waveform translation model, point2wave) using a conditional generative adversarial net (cGAN). Our point2wave is a cGAN pipeline consisting of a generator that translates the waveform corresponding to each point from the input airborne 3-D point cloud and discriminators that calculate the distance between the translated waveform and the ground truth waveform. Using a set of point clouds and waveforms dataset, we have experimented to translate points into the waveforms by point2wave. Experimental results showed that point2wave could translate waveforms from the airborne 3-D point cloud and the translated fake waveforms achieved nearly the same land cover classification performance as the real waveforms.
引用
收藏
页码:11630 / 11642
页数:13
相关论文
共 43 条
[31]   Learning an Attention Model for Robust 2-D/3-D Registration Using Point-To-Plane Correspondences [J].
Schaffert, Roman ;
Wang, Jian ;
Fischer, Peter ;
Borsdorf, Anja ;
Maier, Andreas .
IEEE TRANSACTIONS ON MEDICAL IMAGING, 2020, 39 (10) :3159-3174
[32]   Robust Multi-View 2-D/3-D Registration Using Point-To-Plane Correspondence Model [J].
Schaffert, Roman ;
Wang, Jian ;
Fischer, Peter ;
Maier, Andreas ;
Borsdorf, Anja .
IEEE TRANSACTIONS ON MEDICAL IMAGING, 2020, 39 (01) :161-174
[33]   Pseudo-Reference Point Cloud Quality Measurement Based on Joint 2-D and 3-D Distortion Description [J].
Tu, Renwei ;
Jiang, Gangyi ;
Yu, Mei ;
Zhang, Yun ;
Luo, Ting ;
Zhu, Zhongjie .
IEEE TRANSACTIONS ON INSTRUMENTATION AND MEASUREMENT, 2023, 72
[34]   Enhanced 3-D Urban Scene Reconstruction and Point Cloud Densification Using Gaussian Splatting and Google Earth Imagery [J].
Gao, Kyle ;
Lu, Dening ;
He, Hongjie ;
Xu, Linlin ;
Li, Jonathan ;
Gong, Zheng .
IEEE TRANSACTIONS ON GEOSCIENCE AND REMOTE SENSING, 2025, 63
[35]   An object detection approach for detecting damages in heritage sites using 3-D point clouds and 2-D visual data [J].
Pathak, Rachna ;
Saini, Anil ;
Wadhwa, Arnav ;
Sharma, Himanshu ;
Sangwan, Dhiraj .
JOURNAL OF CULTURAL HERITAGE, 2021, 48 :74-82
[36]   Reconstruction of large-scale anisotropic 3D digital rocks from 2D shale images using generative adversarial network [J].
Chi, Peng ;
Sun, Jianmeng ;
Zhang, Ran ;
Luo, Xin ;
Yan, Weichao .
MARINE AND PETROLEUM GEOLOGY, 2024, 170
[37]   A Pipelined Point Cloud Based Neural Network Processor for 3-D Vision With Large-Scale Max Pooling Layer Prediction [J].
Im, Dongseok ;
Han, Donghyeon ;
Kang, Sanghoon ;
Yoo, Hoi-Jun .
IEEE JOURNAL OF SOLID-STATE CIRCUITS, 2022, 57 (02) :661-670
[38]   Corn Seedling Monitoring Using 3-D Point Cloud Data From Terrestrial Laser Scanning and Registered Camera Data [J].
Xu, Lijun ;
Xu, Teng ;
Li, Xiaolu .
IEEE GEOSCIENCE AND REMOTE SENSING LETTERS, 2020, 17 (01) :137-141
[39]   Low-Latency Visual-Based High-Quality 3-D Reconstruction Using Point Cloud Optimization [J].
Chi, Peng ;
Wang, Zhenmin ;
Liao, Haipeng ;
Li, Ting ;
Zhan, Jinhua ;
Wu, Xiangmiao ;
Tian, Jiyu ;
Zhang, Qin .
IEEE SENSORS JOURNAL, 2023, 23 (17) :20055-20065
[40]   No-Reference 3D Point Cloud Quality Assessment Using Multi-View Projection and Deep Convolutional Neural Network [J].
Bourbia, Salima ;
Karine, Ayoub ;
Chetouani, Aladine ;
El Hassouni, Mohammed ;
Jridi, Maher .
IEEE ACCESS, 2023, 11 :26759-26772