Video detection anomaly via low-rank and sparse decompositions

Author

Lam Tran ; Navasca, C. ; Jiebo Luo

Author_Institution

Univ. of Alabama at Birmingham, Univ. of Rochester, Rochester, NY, USA

fYear

2012

fDate

9-9 Nov. 2012

Firstpage

17

Lastpage

20

Abstract

In this paper, we purpose a method for anomaly detection in surveillance video in a tensor framework. We treat a video as a tensor and utilize a stable PCA to decompose it into two tensors, the first tensor is a low rank tensor that consists of background pixels and the second tensor is a sparse tensor that consists of the foreground pixels. The sparse tensor is then analyzed to detect anomaly. The proposed method is a one-shot framework to determine frames that are anomalous in a video.

Keywords

data mining; principal component analysis; sparse matrices; tensors; video signal processing; video surveillance; background pixels; foreground pixels; low rank tensor; low-rank decompositions; one-shot framework; second tensor; sparse decompositions; sparse tensor; stable PCA; tensor framework; video detection anomalysurveillance video anomaly detection; Educational institutions; Matrix converters; Matrix decomposition; Sparse matrices; Sun; Tensile stress; Vectors; Anomaly detection; Stable PCA; Surveillance Video; Tensor;

fLanguage

English

Publisher

ieee

Conference_Titel

Image Processing Workshop (WNYIPW), 2012 Western New York

Conference_Location

New York, NY

Print_ISBN

978-1-4673-5598-8

Type

conf

DOI

10.1109/WNYIPW.2012.6466649

Filename

6466649