SmartGuard FusionNet: A YOLOv5-Based Multi-Sensor Data Fusion Framework for Superior Weapon Detection in Smart Surveillance Systems

In this paper, we endeavor to substantially advance weapon detection in smart surveillance systems by synergizing the YOLOv5 object detection algorithm with a cutting-edge multi-sensor data fusion framework. This innovative integration aims to harness the precision and speed of YOLOv5, enriched by the depth and breadth of data from multiple sensors, including visual, infrared (IR), and thermal, to adeptly identify weapons in a variety of conditions. Through the development and rigorous evaluation of SmartGuard FusionNet, our approach has been quantitatively assessed across scenarios characterized by optimal lighting, low light, high traffic, and diverse backgrounds. Results demonstrate SmartGuard FusionNet's superior performance, achieving an accuracy of up to 94.2%, precision of 96.8%, recall of 95.6%, and a detection speed of 43 frames per second, significantly surpassing existing surveillance models. These findings not only highlight the framework's unparalleled detection accuracy and efficiency but also its robust adaptability across different environmental challenges. Conclusively, the integration of YOLOv5 with multi-sensor data fusion represents a significant leap forward in smart surveillance technology, offering enhanced capabilities for public safety and security in increasingly complex urban environments.