Naslov
Data fusion in observer networks

Autori
Bodrožić, Ljiljana ; Stipaničev, Darko ; Krstinić, Damir

Vrsta, podvrsta i kategorija rada
Radovi u zbornicima skupova, cjeloviti rad (in extenso), znanstveni

Izvornik
Proceedings of The Second International Workshop on Information Fusion and Dissemination in Wireless Sensor Networks / Yang, L.T. ; Biswas, P.K. ; Hussain, S. - Pisa : Institute of Electrical and Electronics Engineers (IEEE), 2007, 1-6

Skup
International Workshop on Information Fusion and Dissemination in Wireless Sensor Networks (2 ; 2007)

Mjesto i datum
Pisa, Italija, 08.10.2007. - 11.10.2007

Vrsta sudjelovanja
Predavanje

Vrsta recenzije
Međunarodna recenzija

Ključne riječi
sensors; image sensors; observers; multi sensor systems; sensor fusion

Sažetak
Sensor network (SN) and wireless sensor network (WSN) are one of the key technologies in development for 21 century, challenging the imagination of scientists and engineers. A sensor network, deployed directly in an area of interest can provide a large volume of data. Increasing the volume of data, we must create robust algorithms and paradigm for data collecting and utilization. In those cases sensor network with more powerful data processing has to be used. We have introduced such a network using the term the “ observer network” . The observer network is more advanced sensor network inspirited and formally mathematically described using theory of perception and a notation of the observer introduced in 1987. by Bennet, Hoffman and Prakash. The paper main topic is data fusion in observer network. Observer network is based on multi sensor data fusion identifying phenomenon of interest. Data fusion is important on several levels. Two of them are of our particular interest: data fusion in service (information) level, used for network operation failure identification and data fusion in application (knowledge) layer used for final identification of observed phenomenon. As an example, our case study, the advanced forest fire monitoring system will be discussed. Forest fires represent a constant threat to ecological system with possible severe impact on the whole environment. Our particular case study was forest fire protection on Croatian coast and islands, because they have quite enhanced summer forest fire risk. The intelligent forest fire monitoring system called iForestFire was designed using the observer network philosophy and successfully tested during 2005 and 2006 fire seasons. In this paper we will focus primarily on data fusion in intelligent forest fire monitoring system designed as an observer network.

Izvorni jezik
Engleski

Znanstvena područja
Računarstvo

POVEZANOST RADA