Detection of small and camouflaged objects
Keywords:
Small object detection, camouflaged object detection, aerial imagery, computer vision, deep learning, object detection, YOLOv8, YOLOv11, image preprocessing, padding, dataset annotation, automated recognition systemsSynopsis
Detecting small and camouflaged objects in aerial images is challenging, especially when data is acquired from drones. In such images, targets often occupy only a very small portion of the frame and can visually blend in with the surrounding terrain. Image quality is also affected by flight altitude, unstable lighting, motion blur, background noise, and intentional camouflage. Therefore, analyzing such data is complex and time-consuming, significantly increasing the likelihood of missing a target object. Therefore, automatic detection is a highly relevant approach to identifying these objects.
This chapter examines the use of deep learning methods for detecting small and camouflaged objects in aerial photos and videos. The study focuses on YOLO-based detectors, as these models combine good detection quality with high processing speed and are suitable for practical applications. Particular attention is given to the comparison of YOLOv8 and YOLOv11. An experimental study was conducted on an annotated dataset created from publicly available video footage. Two model configurations were trained and evaluated. Image resizing by direct stretching was replaced by adding padding to preserve the proportions of objects within the frame. Additionally, the class structure was simplified, reducing ambiguity during training and increasing classification confidence. These changes were tested alongside the transition from YOLOv8n to YOLOv11s.
The results showed that the improved approach provided more stable detection in complex data and significantly reduced training time. The YOLOv11 model demonstrated the best practical results when working with small targets in complex background conditions. The obtained results confirm that modern architectures based on YOLO family models can be effectively applied to automated data analysis and can serve as a foundation for the further development of intelligent decision support systems.
References
Pan, Yu., Li, L., Qin, J., Chen, J.-J., Gardoni, P. (2024). Unmanned aerial vehicle–human collaboration route planning for intelligent infrastructure inspection. Computer-Aided Civil and Infrastructure Engineering, 39 (14), 2074–2104. https://doi.org/10.1111/mice.13176
Zeng, B., Gao, S., Xu, Y., Zhang, Z., Li, F., Wang, C. (2024). Detection of Military Targets on Ground and Sea by UAVs with Low-Altitude Oblique Perspective. Remote Sensing, 16 (7), 1288. https://doi.org/10.3390/rs16071288
Hwang, K.-S., Ma, J. (2024). Military camouflaged object detection with deep learning using dataset development and combination. The Journal of Defense Modeling and Simulation: Applications, Methodology, Technology, 23 (1), 67–78. https://doi.org/10.1177/15485129241233299
Akyon, F. C., Onur Altinuc, S., Temizel, A. (2022). Slicing Aided Hyper Inference and Fine-Tuning for Small Object Detection. 2022 IEEE International Conference on Image Processing (ICIP), 966–970. https://doi.org/10.1109/icip46576.2022.9897990
Li, Y., Li, Q., Pan, J., Zhou, Y., Zhu, H., Wei, H., Liu, C. (2024). SOD-YOLO: Small-Object-Detection Algorithm Based on Improved YOLOv8 for UAV Images. Remote Sensing, 16 (16), 3057. https://doi.org/10.3390/rs16163057
Han, Y., Guo, J., Yang, H., Guan, R., Zhang, T. (2024). SSMA-YOLO: A Lightweight YOLO Model with Enhanced Feature Extraction and Fusion Capabilities for Drone-Aerial Ship Image Detection. Drones, 8 (4), 145. https://doi.org/10.3390/drones8040145
Yang, C., Shen, Y., Wang, L. (2025). EMFE-YOLO: A Lightweight Small Object Detection Model for UAVs. Sensors, 25 (16), 5200. https://doi.org/10.3390/s25165200
Kaidan, E. (2026). Improved object recognition methods in UAVS based on YOLO. Propylaea of Law and Security, 8, 218–221. https://doi.org/10.32620/pls.2025.8.55
Ali, M. L., Zhang, Z. (2024). The YOLO Framework: A Comprehensive Review of Evolution, Applications, and Benchmarks in Object Detection. Computers, 13 (12), 336. https://doi.org/10.3390/computers13120336
Krytskyi, D., Tkachov, I., Pyvovar, M., Karatanov, S., Krikun, A. (2025). Detecting Objects in Photos and Videos Obtained by Aerial Reconnaissance from UAVs. Integrated Computer Technologies in Mechanical Engineering – 2024. Cham: Springer, 270–283. https://doi.org/10.1007/978-3-031-94845-9_23
Kritsky, D., Kaidan, E., Tkachov, I., Lukin, V. (2025). Automatic detection of hidden objects using uavs: modern neural network approaches. Herald of Khmelnytskyi National University. Technical Sciences, 359 (6.2), 193–204. https://doi.org/10.31891/2307-5732-2025-359-98
Mi, Q., Chao, J., Chen, A., Zhang, K., Lai, J. (2026). YOLO11s-UAV: An Advanced Algorithm for Small Object Detection in UAV Aerial Imagery. Journal of Imaging, 12 (2), 69. https://doi.org/10.3390/jimaging12020069
Makarichev, V., Tsekhmystro, R., Lukin, V., Krytskyi, D. (2025). Performance Improvement of Vehicle and Human Localization and Classification by YOLO Family Networks in Noisy UAV Images. Information, 16 (12), 1087. https://doi.org/10.3390/info16121087
Chaini, C., Jha, V. K., Rajnish, K. (2026). A comparison of single-stage and two-stage based crater detectors on the lunar surface. Earth Science Informatics, 19 (2). https://doi.org/10.1007/s12145-025-02066-7
Gu, H., Wu, J., Huang, H. (2026). CASA-RCNN: A Context-Enhanced and Scale-Adaptive Two-Stage Detector for Dense UAV Aerial Scenes. Drones, 10 (2), 133. https://doi.org/10.3390/drones10020133
Stets, S. (2025). Analysis of accuracy and speed of vehicle detection using neural networks YOLOv8 and YOLOv11. Herald of Khmelnytskyi National University. Technical Sciences, 357 (5.2), 123–130. https://doi.org/10.31891/2307-5732-2025-357-74
Downloads
Pages
132-154
Text Identifier of this Chapter
Chapter 5
Published
June 11, 2026
Copyright (c) 2026 Dmytro Krytskyi, Elvira Kaidan, Artem Chekhovsky
How to Cite
Krytskyi, D., Kaidan, E., & Chekhovsky, A. (2026). Detection of small and camouflaged objects. In V. Lukin, S. Kryvenko, F. Li, S. Abramov, V. Abramova, B. Kovalenko, I. Dohtiev, O. Arkhipov, N. Stojanović, B. Bondžulić, P. Mykhalichenko, T. Cherniavska, B. Cherniavskyi, V. Nadtochii, A. Nadtochyi, M. Korkh, Zghurska О., S. Kasian, K. Nakonechna, … I. Kalinichenko, Pattern recognition in surveillance systems and diagnostics (pp. 132-154). Scientific Route OÜ®. https://doi.org/10.21303/978-9908-845-05-0.ch5
