In this paper, a sequential Bayesian framework is proposed to address the task of joint anomaly detection and tracking for surveillance applications in the presence of clutter. This is achieved by modeling the anomaly as a switching unknown control input which goes into action by modifying the expected dynamics of a target and ceases its activity (becomes nonexistent) under nominal behavior. Random Finite Sets (RFS) make it possible to represent the switching nature of the object anomalous behavior and derive a hybrid Bernoulli filter (HBF) that sequentially updates the joint posterior density of a Bernoulli RFS for the unknown velocity input and the object kinematic state. In addition, the proposed HBF has been customized for maritime anomaly detection by using a piecewise Ornstein-Uhlenbeck (OU) stochastic process as dynamic model of vessels. We illustrate the effectiveness of the proposed filter, implemented in Gaussian-mixture form, and compare its performance in a maritime surveillance example with the Interacting Multiple Model Probabilistic Data Association Filter (IMM - PDAF) for different levels of clutter.

Random Finite Set Tracking for Anomaly Detection in the Presence of Clutter

Millefiori L. M.;Braca P.;
2020

Abstract

In this paper, a sequential Bayesian framework is proposed to address the task of joint anomaly detection and tracking for surveillance applications in the presence of clutter. This is achieved by modeling the anomaly as a switching unknown control input which goes into action by modifying the expected dynamics of a target and ceases its activity (becomes nonexistent) under nominal behavior. Random Finite Sets (RFS) make it possible to represent the switching nature of the object anomalous behavior and derive a hybrid Bernoulli filter (HBF) that sequentially updates the joint posterior density of a Bernoulli RFS for the unknown velocity input and the object kinematic state. In addition, the proposed HBF has been customized for maritime anomaly detection by using a piecewise Ornstein-Uhlenbeck (OU) stochastic process as dynamic model of vessels. We illustrate the effectiveness of the proposed filter, implemented in Gaussian-mixture form, and compare its performance in a maritime surveillance example with the Interacting Multiple Model Probabilistic Data Association Filter (IMM - PDAF) for different levels of clutter.
978-1-7281-8942-0
File in questo prodotto:
Non ci sono file associati a questo prodotto.

I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.

Utilizza questo identificativo per citare o creare un link a questo documento: http://hdl.handle.net/11568/1144005
 Attenzione

Attenzione! I dati visualizzati non sono stati sottoposti a validazione da parte dell'ateneo

Citazioni
  • ???jsp.display-item.citation.pmc??? ND
  • Scopus 2
  • ???jsp.display-item.citation.isi??? ND
social impact