Bringing Intelligence to SAR Missions: A Comprehensive Dataset and Evaluation of YOLO for Human Detection in TIR Images

作     者：Rizk, Mostafa Bayad, Israa 

作者机构：Lebanese Int Univ Sch Engn Beirut 146404 Lebanon Lebanese Univ Fac Sci Beirut 90656 Lebanon 

出 版 物：《IEEE ACCESS》 (IEEE Access)

年 卷 期：2025年第13卷

页      面：17208-17235页

核心收录：

基　　金：The authors would like to thank Zahraa Hadwan for proofreading and editing this article. An earlier version of this paper was presented in part at the 2023 IEEE International Conference on Advances in Biomedical Engineering (ICABME) [DOI: 10.1109/ICABME59496.2023.10293139] 

主　　题：Cameras Imaging YOLO Visualization Accuracy Temperature sensors Deep learning Temperature measurement Temperature Reflection Thermal images object detection deep learning convolutional neural networks human detection search and rescue missions 

摘      要：Effective search and rescue (SAR) missions are critical for locating and assisting injured or missing individuals while optimizing resource allocation and minimizing costs. This work aims to enhance the efficiency of these missions by exploring advanced deep learning techniques for precise and efficient human detection in thermal images. The primary focus of this work is on YOLOv8, the latest version of the You Only Look Once (YOLO) object detection method. The paper also evaluates YOLOv7-Tiny, which is the most streamlined variant derived from YOLOv7. To support the investigation, a novel dataset comprising 17,148 thermal images with 90,882 instances of human subjects representing various conditions and scenarios has been carefully curated. This dataset is used for training and evaluating different variants of YOLOv8 and YOLOv7. The evaluation of the trained models reveals the efficacy of YOLOv8 in detecting humans in thermal images, achieving an average precision rate of 95% with the largest model, YOLOv8x, and an average precision rate of 91% with the smallest model, YOLOv8n. The evaluation of YOLOv7-Tiny shows that it achieves an average precision similar to YOLOv8n, which is 48% lighter in size and more practical choice for real-world deployment. Also, the trained models are deployed on graphical processing units. The tiniest trained model, YOLOv8n, achieves an inference rate of 273.6 frames per second (FPS) while the largest model, YOLOv8, achieves an inference rate of 100.29 FPS. The achieved inference rates along with the achieved detection performances meet with the requirement of fast detection of humans in SAR missions.