Forest fire detection system using landsat satellite images for effective disaster management

Abstract

Forest fires are still the most common terrible danger in forests, and they are an uncontrollable natural occurrence that poses a significant threat to both wildlife and human who live there. According to reports, a total of four to six million wildfires have occurred every year in the last decade around the globe. Early detection of a prospective fire occurrence has the potential to lessen the cause or risk of the fire up to 95%. There are a variety of approaches that may be used to safeguard forest resources against fire. A satellite-based system can be used to identify fires, monitor them, and assess their impact. The benefit of employing a satellite-based system is its wide coverage of any location can cover an area. The majority of satellite-based systems are used to monitor forest fire incidents and analyze charred areas. There are a variety of satellite-based technologies that may be used to detect fires via remote sensing. newlineSeveral early detection algorithms have been introduced to detect the forest fires using LANDSAT images. The remote sensing-based fire detection system contains four basic steps. Initially the collected datasets are given for preprocessing. At this step the image co-registration, error correction and removal of cloud and cloud shadow are performed. Adaptive Median Filter is designed to handle the challenge of removing image noise while also reducing image distortion. In general, AMF processing is used to detect whether the center pixel or median value is affected by impulse noise or not. In the second step, based on region of interest the forest area alone taken for feature extraction. newline