Naslov
Measuring and Controlling Cognitive Process of Visual Attention in Forest Fire Monitoring System

Autori
Šerić, Ljiljana ; Krstinić, Damir ; Bogunović, Pero

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

Izvornik
Proceedings of 2019 IEEE 18th International Conference on COGNITIVE INFORMATICS & COGNITIVE COMPUTING / Soda, P ; Fiorini, R.A. ; Wang, Y. ; Jacobs, G. ; Howard, N. ; Widrow, B. ; Feldman , J. - : Institute of Electrical and Electronics Engineers (IEEE), 2019, 416-421

ISBN
978-1-7281-1419-4

Skup
International Conference on Cognitive Informatics & Cognitive Computing

Mjesto i datum
Milano, Italija, 23.07.2019. - 25.07.2019

Vrsta sudjelovanja
Predavanje

Vrsta recenzije
Međunarodna recenzija

Ključne riječi
Visual attention, Video surveillance, Fire detection, Monitoring, EEG signal

Sažetak
Cognitive process of selecting appropriate information plays important role in perception of hazardous phenomenon. Monitoring and early detection of phenomenon are popular measures used for preventing natural and man-induced hazards. Forest fires management can especially benefit from early fire detection. Forest fire video surveillance improves effectiveness of monitoring, especially if automatic fire detection and alarming is a part of a video surveillance system. However, real benefits of monitoring, early detection and fire prevention can only be achieved with human engagement in decision making about the fire hazard. In this paper we report research and preliminary results on the subject of measuring and controlling visual attention in forest fire video monitoring and surveillance system. The research is designed to discover if an operator’s attention can be controlled in a working environment of monitoring operative center (MOC). Control of the visual attention would enable more rapid detection of fire. With this in mind, we designed and executed two experiments. In first experiment we measured the time taken for fire detection by subjects in several different scenarios. In other experiment we performed measurement of EEG signal of subjects with single electrode instrument during monitoring period and fire detection. Analysis of the results shows that time taken to detect phenomenon depends on scenario of the video stream and that subject’s recognition of fire can be detected with non invasive EEG measurement.

Izvorni jezik
Engleski

Znanstvena područja
Računarstvo

POVEZANOST RADA