Transfer Learning Based High-Precision Multi-Class Object Detection for Real-Time UAV Autonomous Landing via YOLOv8l in Unstructured Scenarios

Notice

This is an unedited manuscript accepted for publication and provided as an Article in Press for early access at the author’s request. The article will undergo copyediting, typesetting, and galley proof review before final publication. Please be aware that errors may be identified during production that could affect the content. All legal disclaimers of the journal apply.

Year : 2026 | Volume : 16 | 02 | Page :
By

Vivek Verma,

Vaibhav Singh,

  1. Independent Researcher, Centre for Advanced Studies, Dr. APJ Abdul KalamTechnical University Lucknow, Uttar Pradesh, Inida
  2. Project Research Scientist-I, Centre for Advanced Studies, Dr. APJ Abdul KalamTechnical University Lucknow, Uttar Pradesh, India

Abstract

A significant challenge for autonomous drone landings in unstructured environments is that of reliably detecting and identifying objects in real-time to ensure safety and accuracy of the landing area. This paper presents a well-founded method for solving this problem using the YOLOv8l object detection framework to detect landing zones, obstacles and people in the relevant vicinity of the landing area. The dataset used for the training of the model contained 720 validation images with unique annotated objects across four classifications, including Vehicle, Unmanned Aerial Platform (UAP), Unmanned Aerial Installation (UAI) and Person, which were used for training over 150 epochs to form the network. The overall precision of the trained model was 98.6%, recall was 96.3%, [email protected] = 98.4% and [email protected]:0.95 = 88.5%. These results demonstrate that YOLOv8l is capable of performing reliably in complex environments. YOLOv8l predicts at a rate of 310ms/image when using the CPU which is sufficient enough to be considered ok for use in a real time context. Results suggest that YOLOv8l can also provide a scalable and accurate means for carrying out autonomous land operations on drones and therefore YOLOv8l may also be integrated into future UAV systems as an operational candidate, with enhanced operational robustness consistently.

Keywords: Transfer Learning, Image Classification, Autonomous Drone, Pretrained Networks, Image Classification Pipeline, YOLOv8l, UAV Landing, Object Detection, Autonomous Aerial Systems

How to cite this article: Vivek Verma, Vaibhav Singh. Transfer Learning Based High-Precision Multi-Class Object Detection for Real-Time UAV Autonomous Landing via YOLOv8l in Unstructured Scenarios. Journal of Aerospace Engineering & Technology. 2026; 16(02):-.
How to cite this URL: Vivek Verma, Vaibhav Singh. Transfer Learning Based High-Precision Multi-Class Object Detection for Real-Time UAV Autonomous Landing via YOLOv8l in Unstructured Scenarios. Journal of Aerospace Engineering & Technology. 2026; 16(02):-. Available from: https://journals.stmjournals.com/joaet/article=2026/view=252127

References

  1. Rahman S, Rony JH, Uddin J, Samad MA. Real-time obstacle detection with YOLOv8 in a WSN using UAV aerial photography. Journal of Imaging. 2023 Oct 10;9(10):216.
  2. Zhang X, Zuo G. Small target detection in UAV view based on improved YOLOv8 algorithm. Scientific Reports. 2025 Jan 2;15(1):421.
  3. Liu C, Meng F, Zhu Z, Zhou L. Object detection of UAV aerial image based on YOLOv8. Frontiers in Computing and Intelligent Systems. 2023;5(3):46-50.
  4. Wu H, Wang W, Wang T, Suzuki S. High-Precision Landing on a Moving Platform Based on Drone Vision Using YOLO Algorithm. Drones. 2025 Mar 29;9(4):261.
  5. Serrano KK, Bandala AA. Yolo-based terrain classification for uav safe landing zone detection. In2023 IEEE Region 10 Symposium (TENSYMP) 2023 Sep 6 (pp. 1-5). IEEE.
  6. Zamri FN, Gunawan TS, Yusoff SH, Alzahrani AA, Bramantoro A, Kartiwi M. Enhanced small drone detection using optimized YOLOv8 with attention mechanisms. IEEE Access. 2024 Jun 28;12:90629-43.
  7. Ming-You M, Shen SE, Yi-Cheng H. Enhancing UAV visual landing recognition with YOLO’s object detection by onboard edge computing. Sensors. 2023;23(21):8999.
  8. Baidya R, Jeong H. Simulation and real-life implementation of UAV autonomous landing system based on object recognition and tracking for safe landing in uncertain environments. Frontiers in Robotics and AI. 2024 Oct 18;11:1450266.
  9. Zhang S, Zhang Y, Dou Y, Kou L, Zuo G, Liu Y, Yang L. YOLO-TME: A UAV landing detection algorithm that is suitable for Polar ice Floe base stations. Scientific Reports. 2025 Dec 19.
  10. Mu L, Zhang Y, Xin J, Zhang Y. Vision-based autonomous landing of a quadrotor UAV on a double-layered nested landing marker. InInternational Conference on Guidance, Navigation and Control 2022 Aug 5 (pp. 2995-3004). Singapore: Springer Nature Singapore.
  11. Kadri MB. Enhancing drone autonomy through cloud integration: a comprehensive software architecture for navigation, visual servoing, and control. PeerJ Computer Science. 2025 Sep 30;11:e3238.
Ahead of Print Subscription Review Article
Volume 16
02
Received 22/07/2026
Accepted 24/07/2026
Published 11/08/2026
Publication Time 20 Days

Login

My IP

PlumX Metrics