AI-Powered Disaster Management System Using Satellite Imagery: A Survey

Abstract

Disaster management is all about time; timely response and an accurate assessment are the basis on which disaster damage may be limited and lives saved. Traditional methods of disaster response rely on human analysis and manual interpretation of satellite images, which are slow and prone to human error. Here, AI can prove to be a technology capable of using ML and DL algorithms to analyze vast quantities of satellite imagery in real time. AI-based systems can work to detect areas affected, assess the severity of the damage, and predict the evolution of disasters for better response and resource allocation. The paper presents recent developments in AI-based disaster management with the assistance of satellite imagery, sketching out major challenges and future research directions.