In today’s rapidly evolving world, ensuring public safety and security has become a major challenge due to the increasing number of unlawful activities such as theft, violence, and unauthorized access. This work proposes an AI Based CCTV Surveillance System that leverages computer vision and deep learning techniques to automate monitoring. The system uses the YOLO (You Only Look Once) model for real-time object detection and identifies suspicious objects. A motion-based analysis approach is also implemented to detect abnormal activities. Upon detection, the system captures image evidence, stores it with timestamps, and sends real-time alerts via Telegram. A Flask-based web dashboard enables efficient monitoring and log management. The proposed system improves detection accuracy, reduces manual effort, and enhances response time, making it suitable for smart surveillance applications.
Sheetal et al. (Mon,) studied this question.