06.2 - Flight Dynamics and Control (UAV related)

DEEP LEARNING VS. TRADITIONAL COMPUTER VISION FOR UAV LANDING SITE DETECTION IN ADVERSE WEATHER

R. Schmidt¹, J. Rüter¹; ¹German Aerospace Center (DLR), Germany

Autonomous UAVs require reliable landing site detection in all weather conditions. This study compares state-of-the-art deep learning and traditional computer vision methods for landing site detection under challenging conditions, aiming to identify the most suitable approach. Algorithms are evaluated based on accuracy, computational cost, and hardware requirements for future UAV applications.

View Paper

DOI: 10.71945/icas2026_0288