DEVELOPMENT OF MODELS AND ALGORITHMS FOR ANOMALY DETECTION FROM DRONE IMAGES

Abstract

The rapid advancement of unmanned aerial vehicles (UAVs) has enabled large-scale image acquisition for various applications, including environmental monitoring, agriculture, infrastructure inspection, and security. However, the vast amount of data generated necessitates automated anomaly recognition methods to identify irregularities efficiently. This article examines the development of models and algorithms for anomaly recognition from drone images. It reviews state-of-the-art approaches, outlines the theoretical basis of anomaly detection, and proposes methodological frameworks integrating machine learning, deep learning, and computer vision techniques.