A dynamic MRF model for foreground detection on range data sequences of rotating multi-beam lidar

Csaba Benedek, Dömötör Molnár, T. Szirányi

Research output: Chapter in Book/Report/Conference proceedingConference contribution

4 Citations (Scopus)

Abstract

In this paper, we propose a probabilistic approach for foreground segmentation in 360°-view-angle range data sequences, recorded by a rotating multi-beam Lidar sensor, which monitors the scene from a fixed position. To ensure real-time operation, we project the irregular point cloud obtained by the Lidar, to a cylinder surface yielding a depth image on a regular lattice, and perform the segmentation in the 2D image domain. Spurious effects resulted by quantification error of the discretized view angle, non-linear position corrections of sensor calibration, and background flickering, in particularly due to motion of vegetation, are significantly decreased by a dynamic MRF model, which describes the background and foreground classes by both spatial and temporal features. Evaluation is performed on real Lidar sequences concerning both video surveillance and traffic monitoring scenarios.

Original languageEnglish
Title of host publicationLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Pages87-96
Number of pages10
Volume7854 LNCS
DOIs
Publication statusPublished - 2013
EventInternational Workshop on Advances in Depth Image Analysis and Applications, WDIA 2012 - Tsukuba, Japan
Duration: Nov 11 2012Nov 11 2012

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume7854 LNCS
ISSN (Print)03029743
ISSN (Electronic)16113349

Other

OtherInternational Workshop on Advances in Depth Image Analysis and Applications, WDIA 2012
CountryJapan
CityTsukuba
Period11/11/1211/11/12

Fingerprint

Lidar
Optical radar
Dynamic models
Rotating
Segmentation
Flickering
Range of data
Angle
Sensor
Video Surveillance
Point Cloud
Probabilistic Approach
Sensors
Vegetation
Telecommunication traffic
Quantification
Irregular
Monitor
Calibration
Traffic

Keywords

  • motion segmentation
  • MRF
  • rotating multi-beam Lidar

ASJC Scopus subject areas

  • Computer Science(all)
  • Theoretical Computer Science

Cite this

Benedek, C., Molnár, D., & Szirányi, T. (2013). A dynamic MRF model for foreground detection on range data sequences of rotating multi-beam lidar. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (Vol. 7854 LNCS, pp. 87-96). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 7854 LNCS). https://doi.org/10.1007/978-3-642-40303-3_10

A dynamic MRF model for foreground detection on range data sequences of rotating multi-beam lidar. / Benedek, Csaba; Molnár, Dömötör; Szirányi, T.

Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics). Vol. 7854 LNCS 2013. p. 87-96 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 7854 LNCS).

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Benedek, C, Molnár, D & Szirányi, T 2013, A dynamic MRF model for foreground detection on range data sequences of rotating multi-beam lidar. in Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics). vol. 7854 LNCS, Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 7854 LNCS, pp. 87-96, International Workshop on Advances in Depth Image Analysis and Applications, WDIA 2012, Tsukuba, Japan, 11/11/12. https://doi.org/10.1007/978-3-642-40303-3_10
Benedek C, Molnár D, Szirányi T. A dynamic MRF model for foreground detection on range data sequences of rotating multi-beam lidar. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics). Vol. 7854 LNCS. 2013. p. 87-96. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). https://doi.org/10.1007/978-3-642-40303-3_10
Benedek, Csaba ; Molnár, Dömötör ; Szirányi, T. / A dynamic MRF model for foreground detection on range data sequences of rotating multi-beam lidar. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics). Vol. 7854 LNCS 2013. pp. 87-96 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).
@inproceedings{bfd5a48337a14f909dc0f7b59f9a2669,
title = "A dynamic MRF model for foreground detection on range data sequences of rotating multi-beam lidar",
abstract = "In this paper, we propose a probabilistic approach for foreground segmentation in 360°-view-angle range data sequences, recorded by a rotating multi-beam Lidar sensor, which monitors the scene from a fixed position. To ensure real-time operation, we project the irregular point cloud obtained by the Lidar, to a cylinder surface yielding a depth image on a regular lattice, and perform the segmentation in the 2D image domain. Spurious effects resulted by quantification error of the discretized view angle, non-linear position corrections of sensor calibration, and background flickering, in particularly due to motion of vegetation, are significantly decreased by a dynamic MRF model, which describes the background and foreground classes by both spatial and temporal features. Evaluation is performed on real Lidar sequences concerning both video surveillance and traffic monitoring scenarios.",
keywords = "motion segmentation, MRF, rotating multi-beam Lidar",
author = "Csaba Benedek and D{\"o}m{\"o}t{\"o}r Moln{\'a}r and T. Szir{\'a}nyi",
year = "2013",
doi = "10.1007/978-3-642-40303-3_10",
language = "English",
isbn = "9783642403026",
volume = "7854 LNCS",
series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",
pages = "87--96",
booktitle = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",

}

TY - GEN

T1 - A dynamic MRF model for foreground detection on range data sequences of rotating multi-beam lidar

AU - Benedek, Csaba

AU - Molnár, Dömötör

AU - Szirányi, T.

PY - 2013

Y1 - 2013

N2 - In this paper, we propose a probabilistic approach for foreground segmentation in 360°-view-angle range data sequences, recorded by a rotating multi-beam Lidar sensor, which monitors the scene from a fixed position. To ensure real-time operation, we project the irregular point cloud obtained by the Lidar, to a cylinder surface yielding a depth image on a regular lattice, and perform the segmentation in the 2D image domain. Spurious effects resulted by quantification error of the discretized view angle, non-linear position corrections of sensor calibration, and background flickering, in particularly due to motion of vegetation, are significantly decreased by a dynamic MRF model, which describes the background and foreground classes by both spatial and temporal features. Evaluation is performed on real Lidar sequences concerning both video surveillance and traffic monitoring scenarios.

AB - In this paper, we propose a probabilistic approach for foreground segmentation in 360°-view-angle range data sequences, recorded by a rotating multi-beam Lidar sensor, which monitors the scene from a fixed position. To ensure real-time operation, we project the irregular point cloud obtained by the Lidar, to a cylinder surface yielding a depth image on a regular lattice, and perform the segmentation in the 2D image domain. Spurious effects resulted by quantification error of the discretized view angle, non-linear position corrections of sensor calibration, and background flickering, in particularly due to motion of vegetation, are significantly decreased by a dynamic MRF model, which describes the background and foreground classes by both spatial and temporal features. Evaluation is performed on real Lidar sequences concerning both video surveillance and traffic monitoring scenarios.

KW - motion segmentation

KW - MRF

KW - rotating multi-beam Lidar

UR - http://www.scopus.com/inward/record.url?scp=84883281937&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84883281937&partnerID=8YFLogxK

U2 - 10.1007/978-3-642-40303-3_10

DO - 10.1007/978-3-642-40303-3_10

M3 - Conference contribution

AN - SCOPUS:84883281937

SN - 9783642403026

VL - 7854 LNCS

T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

SP - 87

EP - 96

BT - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

ER -