Object Matching in Distributed Video Surveillance Systems by LDA-Based Appearance Descriptors

Risultato della ricerca: Chapter

10 Citazioni (Scopus)

Abstract

Establishing correspondences among object instances is still challenging in multi-camera surveillance systems, especially when the cameras’ fields of view are non-overlapping. Spatiotemporal constraints can help in solving the correspondence problem but still leave a wide margin of uncertainty. One way to reduce this uncertainty is to use ap- pearance information about the moving objects in the site. In this paper we present the preliminary results of a new method that can capture salient appearance characteristics at each camera node in the network. A Latent Dirichlet Allocation (LDA) model is created and maintained at each node in the camera network. Each object is encoded in terms of the LDA bag-of-words model for appearance. The encoded appearance is then used to establish probable matching across cameras. Preliminary experiments are conducted on a dataset of 20 individuals and comparison against Madden’s I-MCHR is reported.
Lingua originaleEnglish
Titolo della pubblicazione ospiteImage Analysis and Processing
Pagine547-557
Numero di pagine11
Stato di pubblicazionePublished - 2009

Serie di pubblicazioni

NomeLECTURE NOTES IN COMPUTER SCIENCE

Fingerprint

Video Surveillance
Descriptors
Dirichlet
Camera
Cameras
Correspondence Problem
Uncertainty
Vertex of a graph
Moving Objects
Probable
Field of View
Margin
Surveillance
Correspondence
Object
Model
Experiment
Experiments

All Science Journal Classification (ASJC) codes

  • Theoretical Computer Science
  • Computer Science(all)

Cita questo

La Cascia, M., Lo Presti, L., & Sclaroff, S. (2009). Object Matching in Distributed Video Surveillance Systems by LDA-Based Appearance Descriptors. In Image Analysis and Processing (pagg. 547-557). (LECTURE NOTES IN COMPUTER SCIENCE).

Object Matching in Distributed Video Surveillance Systems by LDA-Based Appearance Descriptors. / La Cascia, Marco; Lo Presti, Liliana; Sclaroff, Stan.

Image Analysis and Processing. 2009. pag. 547-557 (LECTURE NOTES IN COMPUTER SCIENCE).

Risultato della ricerca: Chapter

La Cascia, M, Lo Presti, L & Sclaroff, S 2009, Object Matching in Distributed Video Surveillance Systems by LDA-Based Appearance Descriptors. in Image Analysis and Processing. LECTURE NOTES IN COMPUTER SCIENCE, pagg. 547-557.
La Cascia M, Lo Presti L, Sclaroff S. Object Matching in Distributed Video Surveillance Systems by LDA-Based Appearance Descriptors. In Image Analysis and Processing. 2009. pag. 547-557. (LECTURE NOTES IN COMPUTER SCIENCE).
La Cascia, Marco ; Lo Presti, Liliana ; Sclaroff, Stan. / Object Matching in Distributed Video Surveillance Systems by LDA-Based Appearance Descriptors. Image Analysis and Processing. 2009. pagg. 547-557 (LECTURE NOTES IN COMPUTER SCIENCE).
@inbook{9012a2aba9e8449f96078e0ec1d9d32c,
title = "Object Matching in Distributed Video Surveillance Systems by LDA-Based Appearance Descriptors",
abstract = "Establishing correspondences among object instances is still challenging in multi-camera surveillance systems, especially when the cameras’ fields of view are non-overlapping. Spatiotemporal constraints can help in solving the correspondence problem but still leave a wide margin of uncertainty. One way to reduce this uncertainty is to use ap- pearance information about the moving objects in the site. In this paper we present the preliminary results of a new method that can capture salient appearance characteristics at each camera node in the network. A Latent Dirichlet Allocation (LDA) model is created and maintained at each node in the camera network. Each object is encoded in terms of the LDA bag-of-words model for appearance. The encoded appearance is then used to establish probable matching across cameras. Preliminary experiments are conducted on a dataset of 20 individuals and comparison against Madden’s I-MCHR is reported.",
keywords = "Video surveillance, consistent labelling",
author = "{La Cascia}, Marco and {Lo Presti}, Liliana and Stan Sclaroff",
year = "2009",
language = "English",
series = "LECTURE NOTES IN COMPUTER SCIENCE",
pages = "547--557",
booktitle = "Image Analysis and Processing",

}

TY - CHAP

T1 - Object Matching in Distributed Video Surveillance Systems by LDA-Based Appearance Descriptors

AU - La Cascia, Marco

AU - Lo Presti, Liliana

AU - Sclaroff, Stan

PY - 2009

Y1 - 2009

N2 - Establishing correspondences among object instances is still challenging in multi-camera surveillance systems, especially when the cameras’ fields of view are non-overlapping. Spatiotemporal constraints can help in solving the correspondence problem but still leave a wide margin of uncertainty. One way to reduce this uncertainty is to use ap- pearance information about the moving objects in the site. In this paper we present the preliminary results of a new method that can capture salient appearance characteristics at each camera node in the network. A Latent Dirichlet Allocation (LDA) model is created and maintained at each node in the camera network. Each object is encoded in terms of the LDA bag-of-words model for appearance. The encoded appearance is then used to establish probable matching across cameras. Preliminary experiments are conducted on a dataset of 20 individuals and comparison against Madden’s I-MCHR is reported.

AB - Establishing correspondences among object instances is still challenging in multi-camera surveillance systems, especially when the cameras’ fields of view are non-overlapping. Spatiotemporal constraints can help in solving the correspondence problem but still leave a wide margin of uncertainty. One way to reduce this uncertainty is to use ap- pearance information about the moving objects in the site. In this paper we present the preliminary results of a new method that can capture salient appearance characteristics at each camera node in the network. A Latent Dirichlet Allocation (LDA) model is created and maintained at each node in the camera network. Each object is encoded in terms of the LDA bag-of-words model for appearance. The encoded appearance is then used to establish probable matching across cameras. Preliminary experiments are conducted on a dataset of 20 individuals and comparison against Madden’s I-MCHR is reported.

KW - Video surveillance

KW - consistent labelling

UR - http://hdl.handle.net/10447/54098

M3 - Chapter

T3 - LECTURE NOTES IN COMPUTER SCIENCE

SP - 547

EP - 557

BT - Image Analysis and Processing

ER -