Abstract :

Autonomous Landing Is A Critical Function For Modern Unmanned Aerial Vehicles (UAVs), Especially In GPS-denied, Cluttered, Or Emergency Environments. Traditional Landing Site Detection Approaches Rely Heavily On Handcrafted Features And Domain-specific Rules, Resulting In Limited Adaptability To New Terrains And Lighting Variations. This Paper Proposes A Transfer-learning-based Landing Scene Recognition Framework That Leverages Pretrained Deep Convolutional Neural Networks (CNNs) To Classify And Detect Safe Landing Zones For Drones. By Fine-tuning High-level Semantic Layers Of Established Architectures Such As ResNet And MobileNet, The System Achieves Improved Robustness Against Visual Noise, Occlusion, And Environmental Shifts. Experimental Analysis Demonstrates Significant Improvements In Classification Accuracy And Generalization For Diverse Aerial Imagery Datasets Compared To Conventional Feature-based Systems [1], [4]. The Proposed Method Enhances Drone Autonomy, Supporting Reliable And Context-aware Landing Decisions During Mission-critical Operations [7]. Keywords— Autonomous Landing, UAVs, Transfer Learning, Deep Learning, Landing Scene Recognition, Drone Safety, Aerial Imagery.

Published:

29-11-2025

Issue:

Vol. 25 No. 11 (2025)

Page Nos:

308-314

Section:

Articles

License:

This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License.

How to Cite