Guard Vision: Realtime Weapon Detection

Surveillance systems are necessary for monitoring and maintaining public safety in many different locations. The growing threat of weapons and other dangerous items in everyday life means that there is an increase in demand for more advanced surveillance systems that are capable of accurately identifying these dangerous items. In this research, we are developing a highly advanced weapon detection system using a YOLO-based algorithm and the OpenCV open-source library. The designed system will provide an integration between a CCTV surveillance system and an implementation of YOLO-based weapon detection capabilities. The combination of weapon detection algorithms and the computer vision processing capabilities that are offered by OpenCV allows for the reliable detection of weapons in multiple environments with varying external factors. In order to speed up the time it takes to detect weapons, we will employ graphics processing unit (GPU) acceleration techniques during processing of the YOLO-based detection model.