MUMBAI, India, Sept. 28 -- Intellectual Property India has published a patent application (202641113028 A) filed by Vallurupalli Nageswara Rao Vignana Jyothi Institute Of Engineering And Technology on September 21, 2026, for An Ai-Enabled Distributed Camera System For Real-Time Face Detection, Identification, And Target Tracking.
Inventor includes Dr. Harshavardhan Awari.
The application for the patent was published on September 25, 2026, under issue no. 39/2026.
Abstract: ABSTRACT [0019] A real-time Multi-Camera Target Face Tracking system is presented that detects, recognizes, and follows a selected target face across several non- overlapping camera streams. The approach addresses common difficulties such as occlusion, lighting variation, and identity inconsistency that arise in practical surveillance settings. Face detection is performed with RetinaFace, identity verification relies on ArcFace embeddings matched by cosine similarity, single-camera continuity is maintained by the CSRT tracker, and optional face restoration is provided by GFPGAN for low-quality frames. Cross-camera association is achieved by comparing live embeddings against a stored reference image of the target. A modular pipeline processes multiple video feeds concurrently and presents results through a Flask-based web interface that displays camera identifiers, dwell times, and annotated bounding boxes. The design supports straightforward component replacement and operates effectively on modest hardware, making it suitable for campuses, shopping areas, transport hubs, and smart-city monitoring applications that require continuous awareness of a person of interest.
Disclaimer: Curated by HT Syndication.
