Pendeteksian Senjata Api pada Manusia dalam Situasi Real-Time Menggunakan Model YOLOv4-Tiny

Sulthoni, Dimas Alifta
Thoyyibah, Thoyyibah

Jurnal Ilmu Komputer
Vol 1 No 2 (2023): Jurnal Ilmu Komputer (Edisi Desember 2023)

This research aims to develop a real-time human firearm detection system using the YOLOv4-tiny method. The system is implemented and tested on public security CCTV cameras to enhance responses to potential security threats. The research results indicate that the developed detection system achieves an accuracy level of approximately 95%. Real-time testing successfully detects various types of firearms, including rifles, shotguns, and handguns. This success demonstrates the potential of YOLOv4-tiny as an effective solution for improving public safety with fast and accurate firearm detection. The research makes a significant contribution to security technology development, offering an efficient means to prevent violent incidents and protect communities effectively.