MUMBAI, India, Sept. 14 -- Intellectual Property India has published a patent application (202641105615 A) filed by Saveetha Engineering College on September 02, 2026, for Deep Learning-Powered Smart Security Surveillance And Intrusion Detection Device.
Inventors include Mr. P. Sankar; Mr. Yuvan Sundar S Reg. No; Mr. Mohamed Abrar M Reg. No; and Mr. Vimalraj R Reg. No.
The application for the patent was published on September 11, 2026, under issue no. 37/2026.
Abstract: The present invention relates to a deep learning-powered smart security surveillance and intrusion detection device for automated monitoring of a protected environment. The device comprises an image acquisition unit, processing unit, deep learning inference module, intrusion-zone analysis module, threat classification module, alert generation module, communication module, and storage module. The image acquisition unit captures real-time image and/or video data, which is processed and analyzed by the deep learning inference module to detect and classify persons and/or objects. The intrusion-zone analysis module determines whether a detected entity enters or remains within a predefined protected zone. A threat classification module evaluates parameters including object classification, spatial position, movement trajectory, duration of presence, and predefined security criteria to determine an intrusion status. Upon detection of a potential security threat, the alert generation module generates a local and/or remote notification, while relevant event information is stored for subsequent verification. The device enables intelligent real-time surveillance, automated intrusion detection, reduced false alarms, rapid security response, and adaptable deployment across residential, institutional, industrial, and restricted environments.
Disclaimer: Curated by HT Syndication.
