Close-Combat Weapon Detection in Crisis Zones using YOLOv8

Enhancing security measures in the crisis-affected regions of Cameroon, which have faced prolonged unrest over the years, is critically important. Detecting close-combat weapon within dense crowds can significantly improve surveillance and safety amidst ongoing challenges. These regions require advanced security solutions to address persistent threats faced by the local population. This study develops a detection model based on the You Only Look Once (YOLOv8) architecture to accurately identify and segment sticks and machetes in these crisis-affected areas. By assembling a diverse dataset that captures various scenarios, orientations, and lighting conditions, the model learns to recognize the distinctive features of these objects. By enhancing security measures through advanced technology, this study aims to contribute to ongoing efforts to safeguard communities and restore stability in these troubled areas.