Loading [MathJax]/extensions/tex2jax.js

Global Vehicle Localization by Sequence Analysis Using LiDAR Features Derived by an Autoencoder

Research output: Chapter in book/report/conference proceedingConference contributionResearchpeer review

Authors

Details

Original languageEnglish
Title of host publication2018 IEEE Intelligent Vehicles Symposium, IV 2018
PublisherInstitute of Electrical and Electronics Engineers Inc.
Pages656-661
Number of pages6
ISBN (electronic)9781538644522
ISBN (print)978-1-5386-4453-9
Publication statusPublished - 2018
Event2018 IEEE Intelligent Vehicles Symposium, IV 2018 - Changshu, Suzhou, China
Duration: 26 Sept 201830 Sept 2018

Publication series

NameIEEE Intelligent Vehicles Symposium
ISSN (Print)1931-0587

Abstract

Global vehicle localization is normally done by GNSS sensors. In case of GNSS outages, such as in urban canyons or tunnels, highly automated cars cannot localize without an initial known position. In this paper we propose a method for absolute localization in a city based on only one 2D laser scanner.In out method, localization is done by matching vertical scan lines captured by a 2D laser scanner mounted on the vehicle with scan lines derived from a reference point cloud of the environment. We use a neural network to derive significant features describing the shape of the scan lines. Every scan line of a reference data set is labeled with a specific cluster-id using a k-means algorithm and stored in a reference graph. The same k-means algorithm is used to label the single scan lines of a test drive. The localization is done via a sequence mining approach, where a sequence with a specific length is matched to the position with the highest correlation in the reference sequence.In our experiments we analyze the effect of several parameters, including the number of features and sequence length. The results show that the algorithm performs with an accuracy of about 1.4 m and a completeness of up to 99%. Even if the input scan is represented by only ten features, the results are betten than those obtained by using the whole range scan in the localization step.

ASJC Scopus subject areas

Cite this

Global Vehicle Localization by Sequence Analysis Using LiDAR Features Derived by an Autoencoder. / Schlichting, Alexander; Feuerhake, Udo.
2018 IEEE Intelligent Vehicles Symposium, IV 2018. Institute of Electrical and Electronics Engineers Inc., 2018. p. 656-661 8500358 (IEEE Intelligent Vehicles Symposium).

Research output: Chapter in book/report/conference proceedingConference contributionResearchpeer review

Schlichting, A & Feuerhake, U 2018, Global Vehicle Localization by Sequence Analysis Using LiDAR Features Derived by an Autoencoder. in 2018 IEEE Intelligent Vehicles Symposium, IV 2018., 8500358, IEEE Intelligent Vehicles Symposium, Institute of Electrical and Electronics Engineers Inc., pp. 656-661, 2018 IEEE Intelligent Vehicles Symposium, IV 2018, Changshu, Suzhou, China, 26 Sept 2018. https://doi.org/10.1109/ivs.2018.8500358
Schlichting, A., & Feuerhake, U. (2018). Global Vehicle Localization by Sequence Analysis Using LiDAR Features Derived by an Autoencoder. In 2018 IEEE Intelligent Vehicles Symposium, IV 2018 (pp. 656-661). Article 8500358 (IEEE Intelligent Vehicles Symposium). Institute of Electrical and Electronics Engineers Inc.. https://doi.org/10.1109/ivs.2018.8500358
Schlichting A, Feuerhake U. Global Vehicle Localization by Sequence Analysis Using LiDAR Features Derived by an Autoencoder. In 2018 IEEE Intelligent Vehicles Symposium, IV 2018. Institute of Electrical and Electronics Engineers Inc. 2018. p. 656-661. 8500358. (IEEE Intelligent Vehicles Symposium). doi: 10.1109/ivs.2018.8500358
Schlichting, Alexander ; Feuerhake, Udo. / Global Vehicle Localization by Sequence Analysis Using LiDAR Features Derived by an Autoencoder. 2018 IEEE Intelligent Vehicles Symposium, IV 2018. Institute of Electrical and Electronics Engineers Inc., 2018. pp. 656-661 (IEEE Intelligent Vehicles Symposium).
Download
@inproceedings{da3bda5708fe437da39ab29c75801283,
title = "Global Vehicle Localization by Sequence Analysis Using LiDAR Features Derived by an Autoencoder",
abstract = "Global vehicle localization is normally done by GNSS sensors. In case of GNSS outages, such as in urban canyons or tunnels, highly automated cars cannot localize without an initial known position. In this paper we propose a method for absolute localization in a city based on only one 2D laser scanner.In out method, localization is done by matching vertical scan lines captured by a 2D laser scanner mounted on the vehicle with scan lines derived from a reference point cloud of the environment. We use a neural network to derive significant features describing the shape of the scan lines. Every scan line of a reference data set is labeled with a specific cluster-id using a k-means algorithm and stored in a reference graph. The same k-means algorithm is used to label the single scan lines of a test drive. The localization is done via a sequence mining approach, where a sequence with a specific length is matched to the position with the highest correlation in the reference sequence.In our experiments we analyze the effect of several parameters, including the number of features and sequence length. The results show that the algorithm performs with an accuracy of about 1.4 m and a completeness of up to 99%. Even if the input scan is represented by only ten features, the results are betten than those obtained by using the whole range scan in the localization step.",
author = "Alexander Schlichting and Udo Feuerhake",
note = "Publisher Copyright: {\textcopyright} 2018 IEEE.; 2018 IEEE Intelligent Vehicles Symposium, IV 2018 ; Conference date: 26-09-2018 Through 30-09-2018",
year = "2018",
doi = "10.1109/ivs.2018.8500358",
language = "English",
isbn = "978-1-5386-4453-9",
series = "IEEE Intelligent Vehicles Symposium",
publisher = "Institute of Electrical and Electronics Engineers Inc.",
pages = "656--661",
booktitle = "2018 IEEE Intelligent Vehicles Symposium, IV 2018",
address = "United States",

}

Download

TY - GEN

T1 - Global Vehicle Localization by Sequence Analysis Using LiDAR Features Derived by an Autoencoder

AU - Schlichting, Alexander

AU - Feuerhake, Udo

N1 - Publisher Copyright: © 2018 IEEE.

PY - 2018

Y1 - 2018

N2 - Global vehicle localization is normally done by GNSS sensors. In case of GNSS outages, such as in urban canyons or tunnels, highly automated cars cannot localize without an initial known position. In this paper we propose a method for absolute localization in a city based on only one 2D laser scanner.In out method, localization is done by matching vertical scan lines captured by a 2D laser scanner mounted on the vehicle with scan lines derived from a reference point cloud of the environment. We use a neural network to derive significant features describing the shape of the scan lines. Every scan line of a reference data set is labeled with a specific cluster-id using a k-means algorithm and stored in a reference graph. The same k-means algorithm is used to label the single scan lines of a test drive. The localization is done via a sequence mining approach, where a sequence with a specific length is matched to the position with the highest correlation in the reference sequence.In our experiments we analyze the effect of several parameters, including the number of features and sequence length. The results show that the algorithm performs with an accuracy of about 1.4 m and a completeness of up to 99%. Even if the input scan is represented by only ten features, the results are betten than those obtained by using the whole range scan in the localization step.

AB - Global vehicle localization is normally done by GNSS sensors. In case of GNSS outages, such as in urban canyons or tunnels, highly automated cars cannot localize without an initial known position. In this paper we propose a method for absolute localization in a city based on only one 2D laser scanner.In out method, localization is done by matching vertical scan lines captured by a 2D laser scanner mounted on the vehicle with scan lines derived from a reference point cloud of the environment. We use a neural network to derive significant features describing the shape of the scan lines. Every scan line of a reference data set is labeled with a specific cluster-id using a k-means algorithm and stored in a reference graph. The same k-means algorithm is used to label the single scan lines of a test drive. The localization is done via a sequence mining approach, where a sequence with a specific length is matched to the position with the highest correlation in the reference sequence.In our experiments we analyze the effect of several parameters, including the number of features and sequence length. The results show that the algorithm performs with an accuracy of about 1.4 m and a completeness of up to 99%. Even if the input scan is represented by only ten features, the results are betten than those obtained by using the whole range scan in the localization step.

UR - http://www.scopus.com/inward/record.url?scp=85056775704&partnerID=8YFLogxK

U2 - 10.1109/ivs.2018.8500358

DO - 10.1109/ivs.2018.8500358

M3 - Conference contribution

AN - SCOPUS:85056775704

SN - 978-1-5386-4453-9

T3 - IEEE Intelligent Vehicles Symposium

SP - 656

EP - 661

BT - 2018 IEEE Intelligent Vehicles Symposium, IV 2018

PB - Institute of Electrical and Electronics Engineers Inc.

T2 - 2018 IEEE Intelligent Vehicles Symposium, IV 2018

Y2 - 26 September 2018 through 30 September 2018

ER -

By the same author(s)