A multistage filtering technique to detect hazards on the ground plane

Huiyu Zhou, Andrew M. Wallace, Patrick R. Green

Research output: Contribution to journalArticle

Abstract

We consider the detection of hazards within the ground plane immediately in front of a moving pedestrian. Consecutive views of the scene are acquired by a standard video camera. Using epipolar constraints between views, detected features are matched to compute the camera motion and reconstruct the 3D geometry. Assuming the ground is planar, projective invariance of the cross-ratio and the presence or absence of significant peaks in a Lomb-Scargle periodogram are used in a region-growing technique to label a triangulated mesh as obstructed or unobstructed ground plane. The eventual goal of our approach is to provide an aid to elderly pedestrians in the avoidance of tripping hazards, for example uneven flagstones or kerbs. © 2002 Elsevier Science B.V. All rights reserved.

Original languageEnglish
Pages (from-to)1453-1461
Number of pages9
JournalPattern Recognition Letters
Volume24
Issue number9-10
DOIs
Publication statusPublished - Jun 2003

Fingerprint

Hazards
Video cameras
Invariance
Labels
Cameras
Geometry

Keywords

  • Co-planarity
  • Feature matching
  • Obstacle detection
  • Periodic analysis

Cite this

Zhou, Huiyu ; Wallace, Andrew M. ; Green, Patrick R. / A multistage filtering technique to detect hazards on the ground plane. In: Pattern Recognition Letters. 2003 ; Vol. 24, No. 9-10. pp. 1453-1461.
@article{560ff6e31c014db0802a04d4c0fa2fac,
title = "A multistage filtering technique to detect hazards on the ground plane",
abstract = "We consider the detection of hazards within the ground plane immediately in front of a moving pedestrian. Consecutive views of the scene are acquired by a standard video camera. Using epipolar constraints between views, detected features are matched to compute the camera motion and reconstruct the 3D geometry. Assuming the ground is planar, projective invariance of the cross-ratio and the presence or absence of significant peaks in a Lomb-Scargle periodogram are used in a region-growing technique to label a triangulated mesh as obstructed or unobstructed ground plane. The eventual goal of our approach is to provide an aid to elderly pedestrians in the avoidance of tripping hazards, for example uneven flagstones or kerbs. {\circledC} 2002 Elsevier Science B.V. All rights reserved.",
keywords = "Co-planarity, Feature matching, Obstacle detection, Periodic analysis",
author = "Huiyu Zhou and Wallace, {Andrew M.} and Green, {Patrick R.}",
year = "2003",
month = "6",
doi = "10.1016/S0167-8655(02)00385-9",
language = "English",
volume = "24",
pages = "1453--1461",
journal = "Pattern Recognition Letters",
issn = "0167-8655",
publisher = "Elsevier",
number = "9-10",

}

A multistage filtering technique to detect hazards on the ground plane. / Zhou, Huiyu; Wallace, Andrew M.; Green, Patrick R.

In: Pattern Recognition Letters, Vol. 24, No. 9-10, 06.2003, p. 1453-1461.

Research output: Contribution to journalArticle

TY - JOUR

T1 - A multistage filtering technique to detect hazards on the ground plane

AU - Zhou, Huiyu

AU - Wallace, Andrew M.

AU - Green, Patrick R.

PY - 2003/6

Y1 - 2003/6

N2 - We consider the detection of hazards within the ground plane immediately in front of a moving pedestrian. Consecutive views of the scene are acquired by a standard video camera. Using epipolar constraints between views, detected features are matched to compute the camera motion and reconstruct the 3D geometry. Assuming the ground is planar, projective invariance of the cross-ratio and the presence or absence of significant peaks in a Lomb-Scargle periodogram are used in a region-growing technique to label a triangulated mesh as obstructed or unobstructed ground plane. The eventual goal of our approach is to provide an aid to elderly pedestrians in the avoidance of tripping hazards, for example uneven flagstones or kerbs. © 2002 Elsevier Science B.V. All rights reserved.

AB - We consider the detection of hazards within the ground plane immediately in front of a moving pedestrian. Consecutive views of the scene are acquired by a standard video camera. Using epipolar constraints between views, detected features are matched to compute the camera motion and reconstruct the 3D geometry. Assuming the ground is planar, projective invariance of the cross-ratio and the presence or absence of significant peaks in a Lomb-Scargle periodogram are used in a region-growing technique to label a triangulated mesh as obstructed or unobstructed ground plane. The eventual goal of our approach is to provide an aid to elderly pedestrians in the avoidance of tripping hazards, for example uneven flagstones or kerbs. © 2002 Elsevier Science B.V. All rights reserved.

KW - Co-planarity

KW - Feature matching

KW - Obstacle detection

KW - Periodic analysis

UR - http://www.scopus.com/inward/record.url?scp=0037410693&partnerID=8YFLogxK

U2 - 10.1016/S0167-8655(02)00385-9

DO - 10.1016/S0167-8655(02)00385-9

M3 - Article

VL - 24

SP - 1453

EP - 1461

JO - Pattern Recognition Letters

JF - Pattern Recognition Letters

SN - 0167-8655

IS - 9-10

ER -