Flood detection and mapping through multi-resolution sensor fusion: integrating UAV optical imagery and satellite SAR data
Mulham Fawakherji, Leila Hashemi-Beni
Abstract
Advancements in remote sensing technology have improved flood mapping through the integration of diverse platforms like satellites and UAVs. Data fusion enhances environmental monitoring by addressing challenges such as cloud cover, inundated vegetation, and complex urban landscapes. While most studies have focused on satellite data fusion, combining high-resolution UAV optical imagery with Sentinel-1 SAR data remains underexplored. This research presents a methodology that integrates UAV optical imagery with Sentinel-1 SAR to generate more accurate flood maps. Using machine learning algorithms, the approach effectively extracts inundated and dry vegetation, achieving accuracy up to 97.6% for inundated vegetation and 81.9% for dry vegetation, demonstrating its potential for improving flood mapping precision.