MUMBAI, India, Sept. 28 -- Intellectual Property India has published a patent application (202521015453 A) filed by Rungta International Skills University; Rungta College Of Engineering & Technology; and Rungta College Of Pharmaceutical Sciences And Research on February 22, 2025, for Real-Time Human Action Recognition Crime Prevention System With Video Analysis, Iot, And Anomaly Detection.
Inventors include Shweta Bandhekar; Nupoor Narayan; Pratik Kumar; Shubham Yadav; Ayush Raj Saini; Abhishek Tiwari; Sitasma Karki; Pulkita Dewangan; Shivangi Sharma; T. Livas; and Gulshan Mourya.
The application for the patent was published on September 25, 2026, under issue no. 39/2026.
Abstract: Real-Time Human Action Recognition Crime Prevention System with Video Analysis, IoT, and Anomaly Detection This invention describes a real-time crime prevention system leveraging advanced human action recognition, designed to detect and respond to suspicious behaviors with high precision. The system integrates high-definition video capture, sophisticated feature extraction techniques, and machine learning models, including Convolutional Neural Networks (CNNs) and Recurrent Neural Networks (RNNs), to classify actions such as loitering, unauthorized access, and abrupt movements. A dynamic anomaly detection unit identifies emerging or untrained threat patterns, enhancing adaptability. The system generates real-time alerts through audio alarms, mobile notifications, and centralized dashboards, accompanied by timestamped evidence and analytical reports. Seamlessly integrated with IoT-enabled devices and smart city infrastructures, it ensures scalability and centralized management across diverse environments, including public spaces, critical facilities, and residential complexes. This invention significantly improves surveillance accuracy, reduces response times, and minimizes reliance on manual monitoring, providing a proactive, efficient, and robust solution for modern crime prevention challenges.
Disclaimer: Curated by HT Syndication.
