Houache Noureddine* and Kechar Bouabdellah Pages 3 - 14 ( 12 )
Forest fire disasters have arisen each year due to a number of factors. The main interest of the authorities is to fight against these fires as early as possible with a minimum of damage, by exploiting recent technologies suitable for this field. In this paper, we present the design and the implementation of a forest fire detection system based on the Wireless Multimedia Sensor Networks (WMSN) technology applied to our region (M'sila forest, Oran city - Algeria) using a field experiment testbed with low cost hardware and software. In our previous study, the designed system detects the fire using a mono modal approach (the sensed data was scalar in nature such as the temperature and humidity). In this work, we enhanced this system by collecting, in addition, richer information sources using cameras as data sources (by capturing images) to eliminate the false alarms which present the main weakness of the first system. We call this new system as Multimedia Forest Fire System (M2FS). Field experiments that we have carried out using the testbed under different scenarios by evaluating the image compression, time constraint and energy consumption, allowed us to validate our chosen technology (Arduino mote) for any application (scalar or multimedia), and also revealed the supremacy of the multimodal approach to mitigate efficiently false alarms.
Energy consumption, environmental monitoring, forest fire detection, multimodal approach, testbed, wireless multimedia sensor networks.
Department of Computer Science, Faculty of Exact and Applied Sciences, University of Oran 1 Ahmed Ben Bella, Oran, Department of Computer Science, Faculty of Exact and Applied Sciences, University of Oran 1 Ahmed Ben Bella, Oran