A Review of Deep Learning-Based Semantic Segmentation for Point Cloud

In recent years, the popularity of depth sensors and 3D scanners has led to a rapid development of 3D point clouds. Semantic segmentation of point cloud, as a key step in understanding 3D scenes, has attracted extensive attention of researchers. Recent advances in this topic are dominantly led by de...

Full description

Bibliographic Details
Main Authors: Jiaying Zhang, Xiaoli Zhao, Zheng Chen, Zhejun Lu
Format: Article
Language:English
Published: IEEE 2019-01-01
Series:IEEE Access
Subjects:
Online Access:https://ieeexplore.ieee.org/document/8930503/
_version_ 1818910039532896256
author Jiaying Zhang
Xiaoli Zhao
Zheng Chen
Zhejun Lu
author_facet Jiaying Zhang
Xiaoli Zhao
Zheng Chen
Zhejun Lu
author_sort Jiaying Zhang
collection DOAJ
description In recent years, the popularity of depth sensors and 3D scanners has led to a rapid development of 3D point clouds. Semantic segmentation of point cloud, as a key step in understanding 3D scenes, has attracted extensive attention of researchers. Recent advances in this topic are dominantly led by deep learning-based methods. In this paper, we provide a survey covering various aspects ranging from indirect segmentation to direct segmentation. Firstly, we review methods of indirect segmentation based on multi-views and voxel grids, as well as direct segmentation methods from different perspectives including point ordering, multi-scale, feature fusion and fusion of graph convolutional neural network (GCNN). Then, the common datasets for point cloud segmentation are exposed to help researchers choose which one is the most suitable for their tasks. Following that, we devote a part of the paper to analyze the quantitative results of these methods. Finally, the development trend of point cloud semantic segmentation technology is prospected.
first_indexed 2024-12-19T22:36:28Z
format Article
id doaj.art-2a6ec79d68f24500875b946f590c8550
institution Directory Open Access Journal
issn 2169-3536
language English
last_indexed 2024-12-19T22:36:28Z
publishDate 2019-01-01
publisher IEEE
record_format Article
series IEEE Access
spelling doaj.art-2a6ec79d68f24500875b946f590c85502022-12-21T20:03:11ZengIEEEIEEE Access2169-35362019-01-01717911817913310.1109/ACCESS.2019.29586718930503A Review of Deep Learning-Based Semantic Segmentation for Point CloudJiaying Zhang0https://orcid.org/0000-0002-0465-6176Xiaoli Zhao1https://orcid.org/0000-0002-5859-3211Zheng Chen2https://orcid.org/0000-0001-7192-900XZhejun Lu3https://orcid.org/0000-0001-6617-4124School of Electronic and Electrical Engineering, Shanghai University of Engineering Science, Shanghai, ChinaSchool of Electronic and Electrical Engineering, Shanghai University of Engineering Science, Shanghai, ChinaSchool of Electronic and Electrical Engineering, Shanghai University of Engineering Science, Shanghai, ChinaSchool of Electronic and Electrical Engineering, Shanghai University of Engineering Science, Shanghai, ChinaIn recent years, the popularity of depth sensors and 3D scanners has led to a rapid development of 3D point clouds. Semantic segmentation of point cloud, as a key step in understanding 3D scenes, has attracted extensive attention of researchers. Recent advances in this topic are dominantly led by deep learning-based methods. In this paper, we provide a survey covering various aspects ranging from indirect segmentation to direct segmentation. Firstly, we review methods of indirect segmentation based on multi-views and voxel grids, as well as direct segmentation methods from different perspectives including point ordering, multi-scale, feature fusion and fusion of graph convolutional neural network (GCNN). Then, the common datasets for point cloud segmentation are exposed to help researchers choose which one is the most suitable for their tasks. Following that, we devote a part of the paper to analyze the quantitative results of these methods. Finally, the development trend of point cloud semantic segmentation technology is prospected.https://ieeexplore.ieee.org/document/8930503/3D point cloudsdeep learningfeature fusiongraph convolutional neural networksemantic segmentation
spellingShingle Jiaying Zhang
Xiaoli Zhao
Zheng Chen
Zhejun Lu
A Review of Deep Learning-Based Semantic Segmentation for Point Cloud
IEEE Access
3D point clouds
deep learning
feature fusion
graph convolutional neural network
semantic segmentation
title A Review of Deep Learning-Based Semantic Segmentation for Point Cloud
title_full A Review of Deep Learning-Based Semantic Segmentation for Point Cloud
title_fullStr A Review of Deep Learning-Based Semantic Segmentation for Point Cloud
title_full_unstemmed A Review of Deep Learning-Based Semantic Segmentation for Point Cloud
title_short A Review of Deep Learning-Based Semantic Segmentation for Point Cloud
title_sort review of deep learning based semantic segmentation for point cloud
topic 3D point clouds
deep learning
feature fusion
graph convolutional neural network
semantic segmentation
url https://ieeexplore.ieee.org/document/8930503/
work_keys_str_mv AT jiayingzhang areviewofdeeplearningbasedsemanticsegmentationforpointcloud
AT xiaolizhao areviewofdeeplearningbasedsemanticsegmentationforpointcloud
AT zhengchen areviewofdeeplearningbasedsemanticsegmentationforpointcloud
AT zhejunlu areviewofdeeplearningbasedsemanticsegmentationforpointcloud
AT jiayingzhang reviewofdeeplearningbasedsemanticsegmentationforpointcloud
AT xiaolizhao reviewofdeeplearningbasedsemanticsegmentationforpointcloud
AT zhengchen reviewofdeeplearningbasedsemanticsegmentationforpointcloud
AT zhejunlu reviewofdeeplearningbasedsemanticsegmentationforpointcloud