Implementasi Algoritma YOLOv8, YOLOv9, dan YOLOv10 untuk Deteksi Senjata Tajam Secara Real Time

Abstract

Automatic detection of hazardous objects such as sharp weapons is increasingly important in supporting public security systems. The You Only Look Once (YOLO) algorithm is a widely used approach in Computer Vision due to its ability to detect objects quickly and accurately. This study aims to analyze and compare the performance of the three latest YOLO versions—YOLOv8, YOLOv9, and YOLOv10—in detecting sharp weapons through a web-based on-demand processing application. Evaluation is conducted using a dataset of sharp weapon images under various visual conditions. Each algorithm is assessed using evaluation metrics such as precision, recall, mean Average Precision (mAP), F1-score, and inference speed. The results show that YOLOv9 outperforms the others with a precision of 0.9577 and a mAP@0.5 of 0.9722. This research contributes to identifying the most optimal algorithm for AI-based surveillance systems, particularly in the context of public safety