Background estimation under rapid gain change in thermal imagery

Hulya Yalcin, Robert Collins, Martial Hebert

Research output: Contribution to journalArticle

10 Citations (Scopus)

Abstract

We consider detection of moving ground vehicles in airborne sequences recorded by a thermal sensor with automatic gain control, using an approach that integrates dense optic flow over time to maintain a model of background appearance and a foreground occlusion layer mask. However, the automatic gain control of the thermal sensor introduces rapid changes in intensity that makes this difficult. In this paper we show that an intensity-clipped affine model of sensor gain is sufficient to describe the behavior of our thermal sensor. We develop a method for gain estimation and compensation that uses sparse flow of corner features to compute the affine background scene motion that brings pairs of frames into alignment prior to estimating change in pixel brightness. Dense optic flow and background appearance modeling is then performed on these motion-compensated and brightness-compensated frames. Experimental results demonstrate that the resulting algorithm can segment ground vehicles from thermal airborne video while building a mosaic of the background layer, despite the presence of rapid gain changes.

Original languageEnglish (US)
Pages (from-to)148-161
Number of pages14
JournalComputer Vision and Image Understanding
Volume106
Issue number2-3
DOIs
StatePublished - May 1 2007

Fingerprint

Ground vehicles
Gain control
Sensors
Luminance
Optics
Masks
Pixels
Hot Temperature
Compensation and Redress

All Science Journal Classification (ASJC) codes

  • Software
  • Signal Processing
  • Computer Vision and Pattern Recognition

Cite this

Yalcin, Hulya ; Collins, Robert ; Hebert, Martial. / Background estimation under rapid gain change in thermal imagery. In: Computer Vision and Image Understanding. 2007 ; Vol. 106, No. 2-3. pp. 148-161.
@article{adacc77af0e74dd7a90aa525c3f2a24e,
title = "Background estimation under rapid gain change in thermal imagery",
abstract = "We consider detection of moving ground vehicles in airborne sequences recorded by a thermal sensor with automatic gain control, using an approach that integrates dense optic flow over time to maintain a model of background appearance and a foreground occlusion layer mask. However, the automatic gain control of the thermal sensor introduces rapid changes in intensity that makes this difficult. In this paper we show that an intensity-clipped affine model of sensor gain is sufficient to describe the behavior of our thermal sensor. We develop a method for gain estimation and compensation that uses sparse flow of corner features to compute the affine background scene motion that brings pairs of frames into alignment prior to estimating change in pixel brightness. Dense optic flow and background appearance modeling is then performed on these motion-compensated and brightness-compensated frames. Experimental results demonstrate that the resulting algorithm can segment ground vehicles from thermal airborne video while building a mosaic of the background layer, despite the presence of rapid gain changes.",
author = "Hulya Yalcin and Robert Collins and Martial Hebert",
year = "2007",
month = "5",
day = "1",
doi = "10.1016/j.cviu.2006.10.007",
language = "English (US)",
volume = "106",
pages = "148--161",
journal = "Computer Vision and Image Understanding",
issn = "1077-3142",
publisher = "Academic Press Inc.",
number = "2-3",

}

Background estimation under rapid gain change in thermal imagery. / Yalcin, Hulya; Collins, Robert; Hebert, Martial.

In: Computer Vision and Image Understanding, Vol. 106, No. 2-3, 01.05.2007, p. 148-161.

Research output: Contribution to journalArticle

TY - JOUR

T1 - Background estimation under rapid gain change in thermal imagery

AU - Yalcin, Hulya

AU - Collins, Robert

AU - Hebert, Martial

PY - 2007/5/1

Y1 - 2007/5/1

N2 - We consider detection of moving ground vehicles in airborne sequences recorded by a thermal sensor with automatic gain control, using an approach that integrates dense optic flow over time to maintain a model of background appearance and a foreground occlusion layer mask. However, the automatic gain control of the thermal sensor introduces rapid changes in intensity that makes this difficult. In this paper we show that an intensity-clipped affine model of sensor gain is sufficient to describe the behavior of our thermal sensor. We develop a method for gain estimation and compensation that uses sparse flow of corner features to compute the affine background scene motion that brings pairs of frames into alignment prior to estimating change in pixel brightness. Dense optic flow and background appearance modeling is then performed on these motion-compensated and brightness-compensated frames. Experimental results demonstrate that the resulting algorithm can segment ground vehicles from thermal airborne video while building a mosaic of the background layer, despite the presence of rapid gain changes.

AB - We consider detection of moving ground vehicles in airborne sequences recorded by a thermal sensor with automatic gain control, using an approach that integrates dense optic flow over time to maintain a model of background appearance and a foreground occlusion layer mask. However, the automatic gain control of the thermal sensor introduces rapid changes in intensity that makes this difficult. In this paper we show that an intensity-clipped affine model of sensor gain is sufficient to describe the behavior of our thermal sensor. We develop a method for gain estimation and compensation that uses sparse flow of corner features to compute the affine background scene motion that brings pairs of frames into alignment prior to estimating change in pixel brightness. Dense optic flow and background appearance modeling is then performed on these motion-compensated and brightness-compensated frames. Experimental results demonstrate that the resulting algorithm can segment ground vehicles from thermal airborne video while building a mosaic of the background layer, despite the presence of rapid gain changes.

UR - http://www.scopus.com/inward/record.url?scp=34247571294&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=34247571294&partnerID=8YFLogxK

U2 - 10.1016/j.cviu.2006.10.007

DO - 10.1016/j.cviu.2006.10.007

M3 - Article

AN - SCOPUS:34247571294

VL - 106

SP - 148

EP - 161

JO - Computer Vision and Image Understanding

JF - Computer Vision and Image Understanding

SN - 1077-3142

IS - 2-3

ER -