Semantic segmentation of road furniture in mobile laser scanning data

Fashuai Li, Matti Lehtomäki, S.J. Oude Elberink, G. Vosselman, Antero Kukko, Eetu Puttonen, Yuwei Chen, Juha Hyyppä

Research output: Contribution to journalArticleAcademicpeer-review

1 Citation (Scopus)

Abstract

Road furniture recognition has become a prevalent issue in the past few years because of its great importance in smart cities and autonomous driving. Previous research has especially focussed on pole-like road furniture, such as traffic signs and lamp posts. Published methods have mainly classified road furniture as individual objects. However, most road furniture consists of a combination of classes, such as a traffic sign mounted on a street light pole. To tackle this problem, we propose a framework to interpret road furniture at a more detailed level. Instead of being interpreted as single objects, mobile laser scanning data of road furniture is decomposed in elements individually labelled as poles, and objects attached to them, such as, street lights, traffic signs and traffic lights. In our framework, we first detect road furniture from unorganised mobile laser scanning point clouds. Then detected road furniture is decomposed into poles and attachments (e.g. traffic signs). In the interpretation stage, we extract a set of features to classify the attachments by utilising a knowledge-driven method and four representative types of machine learning classifiers, which are random forest, support vector machine, Gaussian mixture model and naïve Bayes, to explore the optimal method. The designed features are the unary features of attachments and the spatial relations between poles and their attachments. Two experimental test sites in Enschede dataset and Saunalahti dataset were applied, and Saunalahti dataset was collected in two different epochs. In the experimental results, the random forest classifier outperforms the other methods, and the overall accuracy acquired is higher than 80% in Enschede test site and higher than 90% in both Saunalahti epochs. The designed features play an important role in the interpretation of road furniture. The results of two epochs in the same area prove the high reliability of our framework and demonstrate that our method achieves good transferability with an accuracy over 90% through employing the training data of one epoch to test the data in another epoch.

Original languageEnglish
Pages (from-to)98-113
Number of pages16
JournalISPRS journal of photogrammetry and remote sensing
Volume154
Early online date8 Jun 2019
DOIs
Publication statusPublished - Aug 2019

Fingerprint

semantics
Traffic signs
roads
segmentation
Poles
laser
Semantics
road
Scanning
scanning
Lasers
traffic
lasers
poles
time measurement
attachment
Classifiers
luminaires
streets
classifiers

Keywords

  • Decomposition
  • Interpretation
  • Machine learning classifiers
  • Mobile laser scanning
  • Pole-like road furniture
  • ITC-ISI-JOURNAL-ARTICLE

Cite this

Li, Fashuai ; Lehtomäki, Matti ; Oude Elberink, S.J. ; Vosselman, G. ; Kukko, Antero ; Puttonen, Eetu ; Chen, Yuwei ; Hyyppä, Juha. / Semantic segmentation of road furniture in mobile laser scanning data. In: ISPRS journal of photogrammetry and remote sensing. 2019 ; Vol. 154. pp. 98-113.
@article{469bc43f3ac94afbb014f55f777dff6f,
title = "Semantic segmentation of road furniture in mobile laser scanning data",
abstract = "Road furniture recognition has become a prevalent issue in the past few years because of its great importance in smart cities and autonomous driving. Previous research has especially focussed on pole-like road furniture, such as traffic signs and lamp posts. Published methods have mainly classified road furniture as individual objects. However, most road furniture consists of a combination of classes, such as a traffic sign mounted on a street light pole. To tackle this problem, we propose a framework to interpret road furniture at a more detailed level. Instead of being interpreted as single objects, mobile laser scanning data of road furniture is decomposed in elements individually labelled as poles, and objects attached to them, such as, street lights, traffic signs and traffic lights. In our framework, we first detect road furniture from unorganised mobile laser scanning point clouds. Then detected road furniture is decomposed into poles and attachments (e.g. traffic signs). In the interpretation stage, we extract a set of features to classify the attachments by utilising a knowledge-driven method and four representative types of machine learning classifiers, which are random forest, support vector machine, Gaussian mixture model and na{\"i}ve Bayes, to explore the optimal method. The designed features are the unary features of attachments and the spatial relations between poles and their attachments. Two experimental test sites in Enschede dataset and Saunalahti dataset were applied, and Saunalahti dataset was collected in two different epochs. In the experimental results, the random forest classifier outperforms the other methods, and the overall accuracy acquired is higher than 80{\%} in Enschede test site and higher than 90{\%} in both Saunalahti epochs. The designed features play an important role in the interpretation of road furniture. The results of two epochs in the same area prove the high reliability of our framework and demonstrate that our method achieves good transferability with an accuracy over 90{\%} through employing the training data of one epoch to test the data in another epoch.",
keywords = "Decomposition, Interpretation, Machine learning classifiers, Mobile laser scanning, Pole-like road furniture, ITC-ISI-JOURNAL-ARTICLE",
author = "Fashuai Li and Matti Lehtom{\"a}ki and {Oude Elberink}, S.J. and G. Vosselman and Antero Kukko and Eetu Puttonen and Yuwei Chen and Juha Hyypp{\"a}",
year = "2019",
month = "8",
doi = "10.1016/j.isprsjprs.2019.06.001",
language = "English",
volume = "154",
pages = "98--113",
journal = "ISPRS journal of photogrammetry and remote sensing",
issn = "0924-2716",
publisher = "Elsevier",

}

Semantic segmentation of road furniture in mobile laser scanning data. / Li, Fashuai; Lehtomäki, Matti; Oude Elberink, S.J.; Vosselman, G.; Kukko, Antero; Puttonen, Eetu; Chen, Yuwei; Hyyppä, Juha.

In: ISPRS journal of photogrammetry and remote sensing, Vol. 154, 08.2019, p. 98-113.

Research output: Contribution to journalArticleAcademicpeer-review

TY - JOUR

T1 - Semantic segmentation of road furniture in mobile laser scanning data

AU - Li, Fashuai

AU - Lehtomäki, Matti

AU - Oude Elberink, S.J.

AU - Vosselman, G.

AU - Kukko, Antero

AU - Puttonen, Eetu

AU - Chen, Yuwei

AU - Hyyppä, Juha

PY - 2019/8

Y1 - 2019/8

N2 - Road furniture recognition has become a prevalent issue in the past few years because of its great importance in smart cities and autonomous driving. Previous research has especially focussed on pole-like road furniture, such as traffic signs and lamp posts. Published methods have mainly classified road furniture as individual objects. However, most road furniture consists of a combination of classes, such as a traffic sign mounted on a street light pole. To tackle this problem, we propose a framework to interpret road furniture at a more detailed level. Instead of being interpreted as single objects, mobile laser scanning data of road furniture is decomposed in elements individually labelled as poles, and objects attached to them, such as, street lights, traffic signs and traffic lights. In our framework, we first detect road furniture from unorganised mobile laser scanning point clouds. Then detected road furniture is decomposed into poles and attachments (e.g. traffic signs). In the interpretation stage, we extract a set of features to classify the attachments by utilising a knowledge-driven method and four representative types of machine learning classifiers, which are random forest, support vector machine, Gaussian mixture model and naïve Bayes, to explore the optimal method. The designed features are the unary features of attachments and the spatial relations between poles and their attachments. Two experimental test sites in Enschede dataset and Saunalahti dataset were applied, and Saunalahti dataset was collected in two different epochs. In the experimental results, the random forest classifier outperforms the other methods, and the overall accuracy acquired is higher than 80% in Enschede test site and higher than 90% in both Saunalahti epochs. The designed features play an important role in the interpretation of road furniture. The results of two epochs in the same area prove the high reliability of our framework and demonstrate that our method achieves good transferability with an accuracy over 90% through employing the training data of one epoch to test the data in another epoch.

AB - Road furniture recognition has become a prevalent issue in the past few years because of its great importance in smart cities and autonomous driving. Previous research has especially focussed on pole-like road furniture, such as traffic signs and lamp posts. Published methods have mainly classified road furniture as individual objects. However, most road furniture consists of a combination of classes, such as a traffic sign mounted on a street light pole. To tackle this problem, we propose a framework to interpret road furniture at a more detailed level. Instead of being interpreted as single objects, mobile laser scanning data of road furniture is decomposed in elements individually labelled as poles, and objects attached to them, such as, street lights, traffic signs and traffic lights. In our framework, we first detect road furniture from unorganised mobile laser scanning point clouds. Then detected road furniture is decomposed into poles and attachments (e.g. traffic signs). In the interpretation stage, we extract a set of features to classify the attachments by utilising a knowledge-driven method and four representative types of machine learning classifiers, which are random forest, support vector machine, Gaussian mixture model and naïve Bayes, to explore the optimal method. The designed features are the unary features of attachments and the spatial relations between poles and their attachments. Two experimental test sites in Enschede dataset and Saunalahti dataset were applied, and Saunalahti dataset was collected in two different epochs. In the experimental results, the random forest classifier outperforms the other methods, and the overall accuracy acquired is higher than 80% in Enschede test site and higher than 90% in both Saunalahti epochs. The designed features play an important role in the interpretation of road furniture. The results of two epochs in the same area prove the high reliability of our framework and demonstrate that our method achieves good transferability with an accuracy over 90% through employing the training data of one epoch to test the data in another epoch.

KW - Decomposition

KW - Interpretation

KW - Machine learning classifiers

KW - Mobile laser scanning

KW - Pole-like road furniture

KW - ITC-ISI-JOURNAL-ARTICLE

UR - https://ezproxy2.utwente.nl/login?url=https://doi.org/10.1016/j.isprsjprs.2019.06.001

UR - https://ezproxy2.utwente.nl/login?url=https://library.itc.utwente.nl/login/2019/isi/li_sem.pdf

U2 - 10.1016/j.isprsjprs.2019.06.001

DO - 10.1016/j.isprsjprs.2019.06.001

M3 - Article

VL - 154

SP - 98

EP - 113

JO - ISPRS journal of photogrammetry and remote sensing

JF - ISPRS journal of photogrammetry and remote sensing

SN - 0924-2716

ER -