REAL-TIME CROWD AND VIOLENCE DETECTION USING YOLOV8 ALGORITHM
Abstract
The proposed system aims to develop a real-time crowd and violence detection system using the state-of-the-art YOLOv8 algorithm. The system monitors live video feeds to detect sudden increases in crowd density and violent activities, providing real-time alerts for quick intervention. By leveraging the YOLOv8 model's advanced object detection capabilities, the system ensures high accuracy and speed in identifying potential security threats in public areas such as gatherings, events, and surveillance zones. This technology improves public safety by allowing authorities to respond quickly and proactively to incidents. The proposed system works by first collecting and preprocessing data from various sources, followed by training the YOLOv8 model on a curated dataset of diverse crowd and violence scenarios. This model is then integrated with a user-friendly interface that allows security personnel to interact with the system, receive notifications, and take action. Upon detecting an anomaly, the system triggers both an alert sound and an email notification, ensuring immediate attention. The system’s ability to track crowd behavior and identify violent actions provides an enhanced security monitoring solution that is scalable and reliable. By combining real-time object detection, crowd density analysis, and violence detection into one integrated system, this project contributes significantly to the field of public safety and surveillance. This approach ensures that the system can handle diverse public safety challenges in various real-world settings.