ARTIFICIAL INTELLIGENCE IN PHOTOGRAMMETRIC MEASURING AND MODELLING OF LANDSLIDE EVENTS

Abstract

Landslides are among the most devastating natural hazards worldwide, causing significant loss of life, infrastructure damage, and economic disruption. Accurate detection, mapping, and modeling of landslides are essential components of disaster risk reduction and sustainable land management. In the past, finding and studying landslides depended on field mapping and subjective visual interpretation of aerial photography. While these methods established a theoretical foundation for understanding slope dynamics, they were often constrained by limited spatial resolution, high labor costs, and an inability to provide timely insights in rapidly changing landscapes. The contemporary transition toward automated, data-driven methods is characterized by the connecting of high-resolution remote sensing platforms—such as Unmanned Aerial Vehicles (UAVs) and sophisticated satellite constellations—with advanced machine learning (ML) and deep learning (DL) algorithms. This change has made it possible to measure movements of the ground and identify areas at risk for landslides more accurately than ever before. Recent advancements in artificial intelligence (AI), particularly machine learning (ML) and deep learning (DL) techniques, have revolutionized the processing and interpretation of photogrammetric data. This synergy facilitates improved landslide monitoring, early warning, and risk assessment, contributing to more effective emergency reaction. This paper aims to emphasize both the advantages and limitations of these technologies in advancing landslide science and disaster management.