scholarly journals SN Based Forest Fire Detection and Early Warning System

This paper proposes a generic Sensor Network (SN) based forest fire detection and management system, which is scalable and readily deployable for all environments and terrains. SN are being deployed in critical and hazardous areas, for monitoring as well as for collection of useful environmental data for analysis. In these particular contexts, active fire detection and management in forest prove to be challenging, especially in areas which are remote, unapproachable or at the epicenter of such incident. The proposed system caters to the above said problem by allowing a visual representation of status of the sensor nodes (in real-time) through the use of web map system, connected to the strategically deployed nodes based on geography. This deployment is enhanced with the usage of a low power, high range and low data rate wireless protocol, for which the lifetime of the nodes can be further increased through an appropriate scheduling. Moreover, a system with environmental sensors and 3 level hierarchical network covering a total of 4 sq. km has been designed, to investigate the feasibility of the underlying system with the experimental scenarios of fire incident at certain nodes. Software modules providing detection functionalities have been implemented as prototype in the proposed system. The real time system has proved to provide a better visualization and real time tracking of the fire incidents, which in turn facilitates the fire management system as a whole

Author(s):  
Zhentian Jiao ◽  
Youmin Zhang ◽  
Lingxia Mu ◽  
Jing Xin ◽  
Shangbin Jiao ◽  
...  

2003 ◽  
Vol 24 (1) ◽  
pp. 9-22 ◽  
Author(s):  
I. Galindo ◽  
P. López-Pérez ◽  
M. Evangelista-Salazar

The forest is one of the most important wealth of every country. The forest fires destroys the wildlife habitat, damages the environment, affects the climate, spoils the biological properties of the soil, etc. So the forest fire detection is a major issue in the present decade. At the same time the forest fire have to be detected as fast as possible. In the proposed method, a color spatial segmentation, temporal segmentation, global motion compensation, Support Vector Machine (SVM) classifications are used to detect the fire and to segment the fire from the video sequence. The method is implemented over the two real time data sets. The proposed method is most suitable for segmenting fire events over unconstrained videos in real time.


2020 ◽  
Vol 2020 (13) ◽  
pp. 383-387
Author(s):  
Shixiao Wu ◽  
Chengcheng Guo ◽  
Jianfeng Yang

Sign in / Sign up

Export Citation Format

Share Document