Akademik Veri Yönetim Sistemi
32nd IEEE Conference on Signal Processing and Communications Applications, SIU 2024, Mersin, Türkiye, 15 - 18 Mayıs 2024, (Tam Metin Bildiri)
The accurate identification and classification of military vehicles in various and demanding locations is crucial for defense and surveillance missions. Conventional approaches often encounter difficulties when dealing with the complexities of light reflections, camouflage designs, and changing backgrounds, resulting in diminished accuracy and delayed reaction times. This research aims to overcome these obstacles by investigating cutting-edge YOLO-Net, an advanced deep-learning model, for detecting and classifying military vehicles. The proposed YOLOv8 model overcomes the challenges in detecting military vehicles, including the need for high accuracy in various operational sets. A publicly available dataset covering a wide range of military and civilian vehicles in different environments is used to provide a comprehensive analysis of the YOLOv8 model. The results obtained from the experiments indicate that the YOLOv8x model exceeds existing approaches in terms of detection accuracy by 13.6%, making YOLOv8 an indispensable method for automatic military vehicle identification.