Detection and recognition of road markings in panoramic images

C. Li, I.M. Creusen, L. Hazelhoff, P.H.N. With, de

Research output: Chapter in Book/Report/Conference proceedingConference contributionAcademicpeer-review

2 Citations (Scopus)
414 Downloads (Pure)

Abstract

The detection of road markings has many practical applications, such as advanced driver assistance systems, road maintenance and accurate GPS navigation. In this paper we propose an algorithm to detect and recognize road markings from panoramic images. Our algorithm consists of four steps. First, an inverse perspective mapping is applied to the panoramic image, and the potential road markings are segmented based on their intensity difference compared to the surrounding pixels. Second, we extract the distance between the center and the boundary at regular angular steps of each considered potential road marking segment into a feature vector. Third, each segment is classified using a Support Vector Machine (SVM). Finally, by modeling the lane markings, previous false positive detected segments can be rejected based on their orientation and position relative to the lane markings. Our experiments show that the system is capable of recognizing 93%, 95% and 91% of striped line segments, blocks and arrows respectively, as well as 94% of the lane markings.
Original languageEnglish
Title of host publicationVideo Surveillance and Transportation Imaging Applications, San Francisco, California, USA, February 8-12, 2015
PublisherSPIE
Number of pages10
DOIs
Publication statusPublished - 2015
Eventconference; Electronic Imagining; 2015-02-08; 2015-02-12 -
Duration: 8 Feb 201512 Feb 2015

Publication series

NameProceedings of SPIE
Volume9407

Conference

Conferenceconference; Electronic Imagining; 2015-02-08; 2015-02-12
Period8/02/1512/02/15
OtherElectronic Imagining

Fingerprint

Advanced driver assistance systems
Support vector machines
Global positioning system
Navigation
Pixels
Experiments

Cite this

Li, C., Creusen, I. M., Hazelhoff, L., & With, de, P. H. N. (2015). Detection and recognition of road markings in panoramic images. In Video Surveillance and Transportation Imaging Applications, San Francisco, California, USA, February 8-12, 2015 [9407] (Proceedings of SPIE; Vol. 9407). SPIE. https://doi.org/10.1117/12.2081395
Li, C. ; Creusen, I.M. ; Hazelhoff, L. ; With, de, P.H.N. / Detection and recognition of road markings in panoramic images. Video Surveillance and Transportation Imaging Applications, San Francisco, California, USA, February 8-12, 2015. SPIE, 2015. (Proceedings of SPIE).
@inproceedings{fc71cd561029474dac190c4028295d6c,
title = "Detection and recognition of road markings in panoramic images",
abstract = "The detection of road markings has many practical applications, such as advanced driver assistance systems, road maintenance and accurate GPS navigation. In this paper we propose an algorithm to detect and recognize road markings from panoramic images. Our algorithm consists of four steps. First, an inverse perspective mapping is applied to the panoramic image, and the potential road markings are segmented based on their intensity difference compared to the surrounding pixels. Second, we extract the distance between the center and the boundary at regular angular steps of each considered potential road marking segment into a feature vector. Third, each segment is classified using a Support Vector Machine (SVM). Finally, by modeling the lane markings, previous false positive detected segments can be rejected based on their orientation and position relative to the lane markings. Our experiments show that the system is capable of recognizing 93{\%}, 95{\%} and 91{\%} of striped line segments, blocks and arrows respectively, as well as 94{\%} of the lane markings.",
author = "C. Li and I.M. Creusen and L. Hazelhoff and {With, de}, P.H.N.",
year = "2015",
doi = "10.1117/12.2081395",
language = "English",
series = "Proceedings of SPIE",
publisher = "SPIE",
booktitle = "Video Surveillance and Transportation Imaging Applications, San Francisco, California, USA, February 8-12, 2015",
address = "United States",

}

Li, C, Creusen, IM, Hazelhoff, L & With, de, PHN 2015, Detection and recognition of road markings in panoramic images. in Video Surveillance and Transportation Imaging Applications, San Francisco, California, USA, February 8-12, 2015., 9407, Proceedings of SPIE, vol. 9407, SPIE, conference; Electronic Imagining; 2015-02-08; 2015-02-12, 8/02/15. https://doi.org/10.1117/12.2081395

Detection and recognition of road markings in panoramic images. / Li, C.; Creusen, I.M.; Hazelhoff, L.; With, de, P.H.N.

Video Surveillance and Transportation Imaging Applications, San Francisco, California, USA, February 8-12, 2015. SPIE, 2015. 9407 (Proceedings of SPIE; Vol. 9407).

Research output: Chapter in Book/Report/Conference proceedingConference contributionAcademicpeer-review

TY - GEN

T1 - Detection and recognition of road markings in panoramic images

AU - Li, C.

AU - Creusen, I.M.

AU - Hazelhoff, L.

AU - With, de, P.H.N.

PY - 2015

Y1 - 2015

N2 - The detection of road markings has many practical applications, such as advanced driver assistance systems, road maintenance and accurate GPS navigation. In this paper we propose an algorithm to detect and recognize road markings from panoramic images. Our algorithm consists of four steps. First, an inverse perspective mapping is applied to the panoramic image, and the potential road markings are segmented based on their intensity difference compared to the surrounding pixels. Second, we extract the distance between the center and the boundary at regular angular steps of each considered potential road marking segment into a feature vector. Third, each segment is classified using a Support Vector Machine (SVM). Finally, by modeling the lane markings, previous false positive detected segments can be rejected based on their orientation and position relative to the lane markings. Our experiments show that the system is capable of recognizing 93%, 95% and 91% of striped line segments, blocks and arrows respectively, as well as 94% of the lane markings.

AB - The detection of road markings has many practical applications, such as advanced driver assistance systems, road maintenance and accurate GPS navigation. In this paper we propose an algorithm to detect and recognize road markings from panoramic images. Our algorithm consists of four steps. First, an inverse perspective mapping is applied to the panoramic image, and the potential road markings are segmented based on their intensity difference compared to the surrounding pixels. Second, we extract the distance between the center and the boundary at regular angular steps of each considered potential road marking segment into a feature vector. Third, each segment is classified using a Support Vector Machine (SVM). Finally, by modeling the lane markings, previous false positive detected segments can be rejected based on their orientation and position relative to the lane markings. Our experiments show that the system is capable of recognizing 93%, 95% and 91% of striped line segments, blocks and arrows respectively, as well as 94% of the lane markings.

U2 - 10.1117/12.2081395

DO - 10.1117/12.2081395

M3 - Conference contribution

T3 - Proceedings of SPIE

BT - Video Surveillance and Transportation Imaging Applications, San Francisco, California, USA, February 8-12, 2015

PB - SPIE

ER -

Li C, Creusen IM, Hazelhoff L, With, de PHN. Detection and recognition of road markings in panoramic images. In Video Surveillance and Transportation Imaging Applications, San Francisco, California, USA, February 8-12, 2015. SPIE. 2015. 9407. (Proceedings of SPIE). https://doi.org/10.1117/12.2081395