Abstract
The number of forest fires has increased significantly over the past ten years. It indicates that forest area estimates fires are a very urgent task today. The use of satellite-based data simplifies the process of assessing forest fires. The aim is to develop an information tech-nology for automated forest fire contours detection on digital optical satellite datas in con-ditions of non-stationarity and uncertainty based on convolutional neural networks. The most popular tools for forest fire analysis are considered. This work proposed using hotspots to identify all fire and smoke pixels for automated forest fire contour detection. It made it possible to obtain contour polygons of the corresponding areas with various attrib-utes: position, size, etc. The results are tested on Sentinel 2 satellite images of the Ávila re-gion. The proposed method has an overall accuracy of 94.3% for the selection of forest fires.
Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
