Anomaly Detection in Moving-Camera Video Sequences Using Principal Subspace Analysis
Abstract
This study presents a family of algorithms based on sparse decompositions that detect anomalies in video sequences obtained from slow moving cameras. These algorithms start by computing the union of subspaces that best represents all the frames from a reference (anomaly free) video as a low-rank projection plus a sparse residue. Then, they perform a low-rank representation of a target (possibly anomalous) video by taking advantage of both the union of subspaces and the sparse residue computed from the reference video. Such algorithms provide good detection results while at the same time obviating the need for previous video synchronization. However, this is obtained at the cost of a large computational complexity, which hinders their applicability. Another contribution of this paper approaches this problem by using intrinsic properties of the obtained data representation in order to restrict the search space to the most relevant subspaces, providing computational complexity gains of up to two orders of magnitude. The developed algorithms are shown to cope well with videos acquired in challenging scenarios, as verified by the analysis of 59 videos from the VDAO database that comprises videos with abandoned objects in a cluttered industrial scenario.
- Authors:
-
- COPPE/Univ. Federal do Rio de Janeiro, Rio de Janeiro (Brazil)
- North Carolina State Univ., Raleigh, NC (United States)
- Publication Date:
- Research Org.:
- North Carolina State University, Raleigh, NC (United States)
- Sponsoring Org.:
- USDOE National Nuclear Security Administration (NNSA), Office of Nonproliferation and Verification Research and Development (NA-22)
- OSTI Identifier:
- 1438419
- Grant/Contract Number:
- NA0002576
- Resource Type:
- Accepted Manuscript
- Journal Name:
- IEEE Transactions on Circuits and Systems I: Regular Papers
- Additional Journal Information:
- Journal Volume: 65; Journal Issue: 3; Journal ID: ISSN 1549-8328
- Publisher:
- IEEE
- Country of Publication:
- United States
- Language:
- English
- Subject:
- 47 OTHER INSTRUMENTATION; video anomaly detection; sparse representation; object detection; moving camera; subspace recovery
Citation Formats
Thomaz, Lucas A., Jardim, Eric, da Silva, Allan F., da Silva, Eduardo A. B., Netto, Sergio L., and Krim, Hamid. Anomaly Detection in Moving-Camera Video Sequences Using Principal Subspace Analysis. United States: N. p., 2017.
Web. doi:10.1109/TCSI.2017.2758379.
Thomaz, Lucas A., Jardim, Eric, da Silva, Allan F., da Silva, Eduardo A. B., Netto, Sergio L., & Krim, Hamid. Anomaly Detection in Moving-Camera Video Sequences Using Principal Subspace Analysis. United States. https://doi.org/10.1109/TCSI.2017.2758379
Thomaz, Lucas A., Jardim, Eric, da Silva, Allan F., da Silva, Eduardo A. B., Netto, Sergio L., and Krim, Hamid. Mon .
"Anomaly Detection in Moving-Camera Video Sequences Using Principal Subspace Analysis". United States. https://doi.org/10.1109/TCSI.2017.2758379. https://www.osti.gov/servlets/purl/1438419.
@article{osti_1438419,
title = {Anomaly Detection in Moving-Camera Video Sequences Using Principal Subspace Analysis},
author = {Thomaz, Lucas A. and Jardim, Eric and da Silva, Allan F. and da Silva, Eduardo A. B. and Netto, Sergio L. and Krim, Hamid},
abstractNote = {This study presents a family of algorithms based on sparse decompositions that detect anomalies in video sequences obtained from slow moving cameras. These algorithms start by computing the union of subspaces that best represents all the frames from a reference (anomaly free) video as a low-rank projection plus a sparse residue. Then, they perform a low-rank representation of a target (possibly anomalous) video by taking advantage of both the union of subspaces and the sparse residue computed from the reference video. Such algorithms provide good detection results while at the same time obviating the need for previous video synchronization. However, this is obtained at the cost of a large computational complexity, which hinders their applicability. Another contribution of this paper approaches this problem by using intrinsic properties of the obtained data representation in order to restrict the search space to the most relevant subspaces, providing computational complexity gains of up to two orders of magnitude. The developed algorithms are shown to cope well with videos acquired in challenging scenarios, as verified by the analysis of 59 videos from the VDAO database that comprises videos with abandoned objects in a cluttered industrial scenario.},
doi = {10.1109/TCSI.2017.2758379},
journal = {IEEE Transactions on Circuits and Systems I: Regular Papers},
number = 3,
volume = 65,
place = {United States},
year = {Mon Oct 16 00:00:00 EDT 2017},
month = {Mon Oct 16 00:00:00 EDT 2017}
}
Web of Science