Sorry, you need to enable JavaScript to visit this website.

facebooktwittermailshare

OBJECT-ORIENTED ANOMALY DETECTION IN SURVEILLANCE VIDEOS

Abstract: 

Detecting and localizing anomalies in surveillance videos is an ongoing challenge. Most existing methods are patch or trajectory-based, which lack semantic understanding of scenes and may split targets into pieces. To handle this prob-lem, this paper proposes a novel and effective algorithm by incorporating deep object detection and tracking with full utilization of spatial and temporal information. We propose a new dynamic image by fusing both appearance and mo-tion information and feed it into object detection network, which can detect and classify objects precisely even in dim and crowd scenes. Based on the detected objects, we develop an effective and scale-insensitive feature, named histogram variance of optical flow angle (HVOFA), together with mo-tion energy to find abnormal motion patterns. In order to further discover missing anomalies and reduce false detected ones, we conduct a post-processing step with abnormal object tracking. The proposed algorithm outperforms state-of-the-art methods on standard benchmarks.

up
0 users have voted:

Paper Details

Authors:
Xiaodan Li, Weihai Li, Bin Liu, Qiankun Liu, Nenghai Yu
Submitted On:
20 April 2018 - 10:55am
Short Link:
Type:
Poster
Event:
Presenter's Name:
Xiaodan Li
Paper Code:
2394
Document Year:
2018
Cite

Document Files

ICASSP2018-2394.pdf

(19 downloads)

Subscribe

[1] Xiaodan Li, Weihai Li, Bin Liu, Qiankun Liu, Nenghai Yu, "OBJECT-ORIENTED ANOMALY DETECTION IN SURVEILLANCE VIDEOS", IEEE SigPort, 2018. [Online]. Available: http://sigport.org/3108. Accessed: May. 21, 2018.
@article{3108-18,
url = {http://sigport.org/3108},
author = {Xiaodan Li; Weihai Li; Bin Liu; Qiankun Liu; Nenghai Yu },
publisher = {IEEE SigPort},
title = {OBJECT-ORIENTED ANOMALY DETECTION IN SURVEILLANCE VIDEOS},
year = {2018} }
TY - EJOUR
T1 - OBJECT-ORIENTED ANOMALY DETECTION IN SURVEILLANCE VIDEOS
AU - Xiaodan Li; Weihai Li; Bin Liu; Qiankun Liu; Nenghai Yu
PY - 2018
PB - IEEE SigPort
UR - http://sigport.org/3108
ER -
Xiaodan Li, Weihai Li, Bin Liu, Qiankun Liu, Nenghai Yu. (2018). OBJECT-ORIENTED ANOMALY DETECTION IN SURVEILLANCE VIDEOS. IEEE SigPort. http://sigport.org/3108
Xiaodan Li, Weihai Li, Bin Liu, Qiankun Liu, Nenghai Yu, 2018. OBJECT-ORIENTED ANOMALY DETECTION IN SURVEILLANCE VIDEOS. Available at: http://sigport.org/3108.
Xiaodan Li, Weihai Li, Bin Liu, Qiankun Liu, Nenghai Yu. (2018). "OBJECT-ORIENTED ANOMALY DETECTION IN SURVEILLANCE VIDEOS." Web.
1. Xiaodan Li, Weihai Li, Bin Liu, Qiankun Liu, Nenghai Yu. OBJECT-ORIENTED ANOMALY DETECTION IN SURVEILLANCE VIDEOS [Internet]. IEEE SigPort; 2018. Available from : http://sigport.org/3108