MUMBAI, India, Aug. 17 -- Intellectual Property India has published a patent application (202641065109 A) filed by Srm Trp Engineering College on May 23, 2026, for Ai-Based Campus Surveillance System With Real-Time Abnormal Activity Detection And Instant Alert Notification.
Inventors include R. Abinayan; G. Dhanalakshmi; P. Dheebini; and M. Krishnarani.
The application for the patent was published on August 14, 2026, under issue no. 33/2026.
Abstract: In recent years, AI-powered video surveillance has emerged as a critical technology for ensuring campus safety through automated, real-time detection of security threats and abnormal activities. This invention presents an AI-based campus surveillance system that continuously monitors live CCTV video streams from campus locations and automatically detects and responds to abnormal events. CCTV cameras continuously capture live video from campus premises, and individual frames are extracted and preprocessed using OpenCV for deep learning inference. Each frame is passed to the YOLO deep learning model, which performs real-time object detection and activity classification to identify human presence and detect abnormal events such as intrusion, violence, and falls based on trained patterns. Confidence scores are evaluated to confirm valid detections and reduce false alarms. Upon detection of an abnormal event, the system captures the frame as visual evidence, generates alert details including event type, timestamp, and camera location, and instantly delivers notifications with the captured image to authorized security personnel via a Telegram bot. All event data is simultaneously logged in a database for monitoring, historical analysis, and audit trail maintenance. The proposed system provides a comprehensive, scalable, and efficient solution for intelligent campus security management, combining real-time AI-driven detection, automated evidence capture, instant multi-channel notification, and persistent event logging within a unified surveillance framework.
Disclaimer: Curated by HT Syndication.
