Detect and Alert Unauthorized Weapon Usage with Haar Cascade and YOLOv7

This paper proposes an intelligent security system integrating YOLOv7 and Haar Cascade algorithms for real-time weapon detection and facial recognition in public spaces. The system, equipped with an email alert and alarm sound feature, aims to address the increasing need for advanced security due to rising violent crimes. Utilizing YOLOv7 for object detection and a deep CNN for weapon identification, the system targets applications such as airport security, public transportation, and events to ensure public safety. Identified weapon classes include Handgun, Sword, SMG, Sniper, Automatic Rifle, Bazooka, Grenade Launcher, Knife, and Shotgun. The incorporation of facial recognition enhances suspect identification in real-time, demonstrating the potential to prevent threats in public spaces.