Abstract

Many surveillance and security monitoring videos are long and of low quality. Moreover, reviewing and extracting anomaly events in the videos is a lengthy and manually intensive process. In this paper, we present two efficient anomaly detection algorithms based on saliency to detect anomalous events in low quality videos. The events' start times and durations are saved in a video summary for later reviews. The video summary is very short. For example, we have summarized a 14-minute long video into a 16-second video summary. Extensive evaluations of the two algorithms clearly demonstrated the feasibility of these algorithms. A user friendly software tool has also been developed to help human operators review and confirm those events.

Original languageEnglish (US)
Title of host publicationPattern Recognition and Tracking XXIX
PublisherSPIE
Volume10649
ISBN (Electronic)9781510618091
DOIs
StatePublished - Jan 1 2018
EventPattern Recognition and Tracking XXIX 2018 - Orlando, United States
Duration: Apr 18 2018Apr 19 2018

Other

OtherPattern Recognition and Tracking XXIX 2018
CountryUnited States
CityOrlando
Period4/18/184/19/18

Fingerprint

Video Quality
Anomaly Detection
anomalies
Saliency
software development tools
reviewing
surveillance
Software Tools
Surveillance
Anomalous
Anomaly
Monitoring
operators
evaluation
Evaluation
Operator
Review

Keywords

  • Anomaly detection
  • low quality videos
  • video summarization

ASJC Scopus subject areas

  • Electronic, Optical and Magnetic Materials
  • Condensed Matter Physics
  • Computer Science Applications
  • Applied Mathematics
  • Electrical and Electronic Engineering

Cite this

Kwan, C., Zhou, J., Wang, Z., & Li, B. (2018). Efficient anomaly detection algorithms for summarizing low quality videos. In Pattern Recognition and Tracking XXIX (Vol. 10649). [1064906] SPIE. https://doi.org/10.1117/12.2303764

Efficient anomaly detection algorithms for summarizing low quality videos. / Kwan, Chiman; Zhou, Jin; Wang, Zheshen; Li, Baoxin.

Pattern Recognition and Tracking XXIX. Vol. 10649 SPIE, 2018. 1064906.

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

Kwan, C, Zhou, J, Wang, Z & Li, B 2018, Efficient anomaly detection algorithms for summarizing low quality videos. in Pattern Recognition and Tracking XXIX. vol. 10649, 1064906, SPIE, Pattern Recognition and Tracking XXIX 2018, Orlando, United States, 4/18/18. https://doi.org/10.1117/12.2303764
Kwan C, Zhou J, Wang Z, Li B. Efficient anomaly detection algorithms for summarizing low quality videos. In Pattern Recognition and Tracking XXIX. Vol. 10649. SPIE. 2018. 1064906 https://doi.org/10.1117/12.2303764
Kwan, Chiman ; Zhou, Jin ; Wang, Zheshen ; Li, Baoxin. / Efficient anomaly detection algorithms for summarizing low quality videos. Pattern Recognition and Tracking XXIX. Vol. 10649 SPIE, 2018.
@inproceedings{09fe664d68ee40cdba0b0d486fd158fe,
title = "Efficient anomaly detection algorithms for summarizing low quality videos",
abstract = "Many surveillance and security monitoring videos are long and of low quality. Moreover, reviewing and extracting anomaly events in the videos is a lengthy and manually intensive process. In this paper, we present two efficient anomaly detection algorithms based on saliency to detect anomalous events in low quality videos. The events' start times and durations are saved in a video summary for later reviews. The video summary is very short. For example, we have summarized a 14-minute long video into a 16-second video summary. Extensive evaluations of the two algorithms clearly demonstrated the feasibility of these algorithms. A user friendly software tool has also been developed to help human operators review and confirm those events.",
keywords = "Anomaly detection, low quality videos, video summarization",
author = "Chiman Kwan and Jin Zhou and Zheshen Wang and Baoxin Li",
year = "2018",
month = "1",
day = "1",
doi = "10.1117/12.2303764",
language = "English (US)",
volume = "10649",
booktitle = "Pattern Recognition and Tracking XXIX",
publisher = "SPIE",

}

TY - GEN

T1 - Efficient anomaly detection algorithms for summarizing low quality videos

AU - Kwan, Chiman

AU - Zhou, Jin

AU - Wang, Zheshen

AU - Li, Baoxin

PY - 2018/1/1

Y1 - 2018/1/1

N2 - Many surveillance and security monitoring videos are long and of low quality. Moreover, reviewing and extracting anomaly events in the videos is a lengthy and manually intensive process. In this paper, we present two efficient anomaly detection algorithms based on saliency to detect anomalous events in low quality videos. The events' start times and durations are saved in a video summary for later reviews. The video summary is very short. For example, we have summarized a 14-minute long video into a 16-second video summary. Extensive evaluations of the two algorithms clearly demonstrated the feasibility of these algorithms. A user friendly software tool has also been developed to help human operators review and confirm those events.

AB - Many surveillance and security monitoring videos are long and of low quality. Moreover, reviewing and extracting anomaly events in the videos is a lengthy and manually intensive process. In this paper, we present two efficient anomaly detection algorithms based on saliency to detect anomalous events in low quality videos. The events' start times and durations are saved in a video summary for later reviews. The video summary is very short. For example, we have summarized a 14-minute long video into a 16-second video summary. Extensive evaluations of the two algorithms clearly demonstrated the feasibility of these algorithms. A user friendly software tool has also been developed to help human operators review and confirm those events.

KW - Anomaly detection

KW - low quality videos

KW - video summarization

UR - http://www.scopus.com/inward/record.url?scp=85049221795&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85049221795&partnerID=8YFLogxK

U2 - 10.1117/12.2303764

DO - 10.1117/12.2303764

M3 - Conference contribution

AN - SCOPUS:85049221795

VL - 10649

BT - Pattern Recognition and Tracking XXIX

PB - SPIE

ER -