Dynamic data driven sensor array fusion for target detection and classification

Nurali Virani, Shane Marcks, Soumalya Sarkar, Kushal Mukherjee, Asok Ray, Shashi Phoha

Research output: Contribution to journalConference article

11 Citations (Scopus)

Abstract

Target detection and classification using unattended ground sensors (UGS) has been addressed in literature. Various techniques have been proposed for target detection, but target classification is a challenging task to accomplish using the limited processing power on each sensor module. The major hindrance in using these sensors reliably is, that, the sensor observations are significantly affected by external conditions, which are referred to as context. When the context is slowly time-varying (e.g., day-night cycling and seasonal variations) the usage of the same classifier may not be a good way to perform target classification. In this paper, a new framework is proposed as a Dynamic Data Driven Application System (DDDAS) to dynamically extract and use the knowledge of context as feedback in order to adaptively choose the appropriate classifiers and thereby enhance the target classification performance. The features are extracted by symbolic dynamic filtering (SDF) from the time series of sensors in an array and spatio-temporal aggregation of these features represents the context. Then, a context evolution model is constructed as a deterministic finite state automata (DFSA) and, for every context state in this DFSA, an event classifier is trained to classify the targets. The proposed technique of detection and classification has been compared with a traditional method of training classifiers without using any contextual information.

Original languageEnglish (US)
Pages (from-to)2046-2055
Number of pages10
JournalProcedia Computer Science
Volume18
DOIs
StatePublished - Jan 1 2013
Event13th Annual International Conference on Computational Science, ICCS 2013 - Barcelona, Spain
Duration: Jun 5 2013Jun 7 2013

Fingerprint

Sensor arrays
Target tracking
Fusion reactions
Classifiers
Sensors
Finite automata
Time series
Agglomeration
Feedback
Processing

All Science Journal Classification (ASJC) codes

  • Computer Science(all)

Cite this

Virani, Nurali ; Marcks, Shane ; Sarkar, Soumalya ; Mukherjee, Kushal ; Ray, Asok ; Phoha, Shashi. / Dynamic data driven sensor array fusion for target detection and classification. In: Procedia Computer Science. 2013 ; Vol. 18. pp. 2046-2055.
@article{0d83829893b14eee9a00fdcc612e26b0,
title = "Dynamic data driven sensor array fusion for target detection and classification",
abstract = "Target detection and classification using unattended ground sensors (UGS) has been addressed in literature. Various techniques have been proposed for target detection, but target classification is a challenging task to accomplish using the limited processing power on each sensor module. The major hindrance in using these sensors reliably is, that, the sensor observations are significantly affected by external conditions, which are referred to as context. When the context is slowly time-varying (e.g., day-night cycling and seasonal variations) the usage of the same classifier may not be a good way to perform target classification. In this paper, a new framework is proposed as a Dynamic Data Driven Application System (DDDAS) to dynamically extract and use the knowledge of context as feedback in order to adaptively choose the appropriate classifiers and thereby enhance the target classification performance. The features are extracted by symbolic dynamic filtering (SDF) from the time series of sensors in an array and spatio-temporal aggregation of these features represents the context. Then, a context evolution model is constructed as a deterministic finite state automata (DFSA) and, for every context state in this DFSA, an event classifier is trained to classify the targets. The proposed technique of detection and classification has been compared with a traditional method of training classifiers without using any contextual information.",
author = "Nurali Virani and Shane Marcks and Soumalya Sarkar and Kushal Mukherjee and Asok Ray and Shashi Phoha",
year = "2013",
month = "1",
day = "1",
doi = "10.1016/j.procs.2013.05.374",
language = "English (US)",
volume = "18",
pages = "2046--2055",
journal = "Procedia Computer Science",
issn = "1877-0509",
publisher = "Elsevier BV",

}

Dynamic data driven sensor array fusion for target detection and classification. / Virani, Nurali; Marcks, Shane; Sarkar, Soumalya; Mukherjee, Kushal; Ray, Asok; Phoha, Shashi.

In: Procedia Computer Science, Vol. 18, 01.01.2013, p. 2046-2055.

Research output: Contribution to journalConference article

TY - JOUR

T1 - Dynamic data driven sensor array fusion for target detection and classification

AU - Virani, Nurali

AU - Marcks, Shane

AU - Sarkar, Soumalya

AU - Mukherjee, Kushal

AU - Ray, Asok

AU - Phoha, Shashi

PY - 2013/1/1

Y1 - 2013/1/1

N2 - Target detection and classification using unattended ground sensors (UGS) has been addressed in literature. Various techniques have been proposed for target detection, but target classification is a challenging task to accomplish using the limited processing power on each sensor module. The major hindrance in using these sensors reliably is, that, the sensor observations are significantly affected by external conditions, which are referred to as context. When the context is slowly time-varying (e.g., day-night cycling and seasonal variations) the usage of the same classifier may not be a good way to perform target classification. In this paper, a new framework is proposed as a Dynamic Data Driven Application System (DDDAS) to dynamically extract and use the knowledge of context as feedback in order to adaptively choose the appropriate classifiers and thereby enhance the target classification performance. The features are extracted by symbolic dynamic filtering (SDF) from the time series of sensors in an array and spatio-temporal aggregation of these features represents the context. Then, a context evolution model is constructed as a deterministic finite state automata (DFSA) and, for every context state in this DFSA, an event classifier is trained to classify the targets. The proposed technique of detection and classification has been compared with a traditional method of training classifiers without using any contextual information.

AB - Target detection and classification using unattended ground sensors (UGS) has been addressed in literature. Various techniques have been proposed for target detection, but target classification is a challenging task to accomplish using the limited processing power on each sensor module. The major hindrance in using these sensors reliably is, that, the sensor observations are significantly affected by external conditions, which are referred to as context. When the context is slowly time-varying (e.g., day-night cycling and seasonal variations) the usage of the same classifier may not be a good way to perform target classification. In this paper, a new framework is proposed as a Dynamic Data Driven Application System (DDDAS) to dynamically extract and use the knowledge of context as feedback in order to adaptively choose the appropriate classifiers and thereby enhance the target classification performance. The features are extracted by symbolic dynamic filtering (SDF) from the time series of sensors in an array and spatio-temporal aggregation of these features represents the context. Then, a context evolution model is constructed as a deterministic finite state automata (DFSA) and, for every context state in this DFSA, an event classifier is trained to classify the targets. The proposed technique of detection and classification has been compared with a traditional method of training classifiers without using any contextual information.

UR - http://www.scopus.com/inward/record.url?scp=84896980341&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84896980341&partnerID=8YFLogxK

U2 - 10.1016/j.procs.2013.05.374

DO - 10.1016/j.procs.2013.05.374

M3 - Conference article

VL - 18

SP - 2046

EP - 2055

JO - Procedia Computer Science

JF - Procedia Computer Science

SN - 1877-0509

ER -