Detecting The Small Object Recognition By Drone Images Using Yolo

Abstract

Unmanned Aerial Vehicle (UAV) imagery has become an indispensable resource for applications including traffic surveillance, disaster management, and airspace monitoring due to its versatility, portability, and cost-effectiveness. Nonetheless, object detection in UAV images remains a challenging task, largely because of small object scales, complex and cluttered backgrounds, and high noise levels. This study introduces a novel object detection framework based on YOLOv10, a state-of-the-art model renowned for its efficient architecture and superior detection accuracy. The proposed approach is specifically tailored for UAV aerial imagery, emphasizing the enhancement of small object detection through advanced feature extraction and improved spatial reasoning. By incorporating adaptive feature enhancement and deep semantic learning techniques, the model achieves robust performance even under challenging imaging conditions. Moreover, leveraging convolutional attention mechanisms, multi-scale detection heads, and optimized backbone architectures enables the system to capture fine-grained details while sustaining real-time processing capabilities. Experimental results demonstrate that the YOLOv10-based framework provides accurate and reliable object detection in complex UAV scenarios, highlighting its potential as a powerful tool for aerial image analysis.