YOLO-Based Weapon Detection and Mail Alert System Using Machine Learning
M Rama Raju,
K Bhavani,
MD Mudasir,
M Rahul,
M Tejasri
This paper presents the design and development of a YOLO-Based Weapon Detection and Mail Alert
System aimed at enhancing real-time surveillance and public safety. The system leverages deep learning
techniques, specifically the YOLO (You Only Look Once) object detection algorithm, to identify weapons
such as guns and knives from live video streams or uploaded images.
The proposed system integrates computer vision using OpenCV, real-time detection using YOLOv8,
and automated email alert mechanisms using SMTP. Upon detecting a weapon, the system captures the
frame, highlights the detected object, and sends an instant alert email with the image, timestamp, and
location details to authorized personnel.
The system is implemented using Python and Streamlit, providing an interactive interface for users.
Experimental results show that the system achieves high detection accuracy, fast response time, and
reduced dependency on manual surveillance. This approach significantly improves security monitoring
efficiency and enables proactive threat detection in real-world environments.
