Naslov
Wildfire video observers network with physical and virtual sensors

Autori
Stipaničev, Darko ; Šerić, Ljiljana ; Krstinić, Damir ; Bugarić, Marin

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

Izvornik
Proceeding of 10th EARSeL Forest Fire Special Interest Group Workshop - Sensors, Multi-Sensor Integration, large Volumes: New opportunities and Challanges in Forest Fire Research / Themistocleous, Kyriacos ; Hadjimitsis, Diofantos ; Gitas, Ioannios ; Boschetti, Luigi - Limassol, 2015

ISBN
978-9963-697-15-1

Skup
10th EARSeL Forest Fire Special Interest Group Workshop - Sensors, Multi-Sensor Integration, large Volumes: New opportunities and Challanges in Forest Fire Research

Mjesto i datum
Limassol, Cipar, 02.11.2015. - 05.11.2015

Vrsta sudjelovanja
Predavanje

Vrsta recenzije
Međunarodna recenzija

Ključne riječi
forest fires; wildfires; smoke and fire detection; wildfire monitoring; wildfire surveillance; video based wildfire monitoring network; fusion of physical and virtual sensors

Sažetak
Wildfires still represent a constant threat to ecological systems, infrastructure and human lives, particularly in Mediterranean area. Apart from preventive measures, early fire detection and quick and appropriate intervention are the most important measures for wildfire damage minimization. Once a wildfire has expanded it is more difficult to control it and extinguish. Therefore in countries that have a great wildfire threat a lot of effort is done to organize appropriate and efficient wildfire surveillance and monitoring system. Traditionally wildfire surveillance was based on human observers located on selected locations in appropriate monitoring towers. Human observers primarily use their sight as an observation sensor trying to spot the wildfire smoke, because in the smoke is visible long before the flame, particularly in environments with forests and dense vegetation. The main task of human wildfire observer is to detect the wildfire in incipient stage, try to locate it on the map and alerts the appropriate fire department. Technically and functionally better wildfire monitoring system could be realized as video cameras based human wildfire surveillance and monitoring. In this case remotely controlled video cameras are installed on monitoring spot and the human observers are located in the observation center. Such a system could be used not only for early fire detection, but also for distant video presence. If the wildfire-monitoring network conceived of multiple cameras covering the whole region is established, the distant video presence is even more important, particularly if the monitored location could be seen from more then one monitoring spots. The main limitation of video cameras based human surveillance is that fire detection depends entirely on the human observation. The observer is located in more comfortable environment, the observation center, but he or she has to watch carefully multiple computer monitors all the time, so problems like fatigue, boredom and lost of concentration could be encountered. That was the main reason for introduction of various forms of automatic and advanced automatic wildfire video camera based surveillance and monitoring systems and networks. One unit of wildfire video based wildfire surveillance system could be threated as video based wildfire observer. A good theoretical background for wildfire observer could be a formal theory of perception and notation of observer as a unit capable to detect observation phenomena (wildfire smoke) using physical sensors (video camera, meteorological station), as well as virtual sensors (day-night sensor, false alarm reduction sensors etc.). Aristotle coined that “The whole is more than the sum of its parts.” This sentence was later used to describe the synergy, the effect that exists between parts of a system working together in a cooperative effort. Therefore, having a video based wildfire observance network of mutually connected individual wildfire observers working together in cooperation, could give the overall effect greater than the sum of their individual effects. Cooperation of individual wildfire observers could be used to improve automatic wildfire detection as well as to enhance manual cameras PTZ control. In this paper an example of advanced video based automatic wildfire surveillance and monitoring network developed at University of Split is described. FESB video based wildfire observer network is not just another application of well-known wildfire video monitoring technology. It is innovative for its theoretical foundations based on observer network theory and formal theory of perception and its architecture and implementation, having a lot of new features not found in conventional wildfire video monitoring systems. The system is both integral and intelligent. It is integral because it is based on fusion of different data types and fusion of physical and virtual sensors. It is intelligent because it is based on various artificial intelligence technologies from multi-agent organization to advance detection algorithms ; alarm reduction techniques and augmented reality based multiple cameras control capabilities.

Izvorni jezik
Engleski

Znanstvena područja
Računarstvo

POVEZANOST RADA