Software Development for the Wildfires Occurrence Detection Based on an Ensemble of Algorithms for Satellite Data Processing
Year: 2023
Type: conference-paper
Abstract: In the presented work, it is proposed to implement satellite imagery and automatic wildfires occurrence detection based on data from the MODIS/Terra (Aqua) and AVHRR/NOAA scanning systems. The software development uses a conconvolutional neural network trained on a set of image patches and corresponding segmentation masks obtained on the basis of the most well-known methods for detecting active fires. The decision to fix the wildfires occurrence is made based on the use of a voting scheme: the results of at least two algorithms must match.
Authors: I. O. Pochinok, I. M. Lazareva, O. I. Lyash
Language: English
Subfield: Safety, Risk, Reliability and Quality
Field: Engineering
Domain: Physical Sciences
Open Access status: gold
OpenAlex ID: W4381294872