MUMBAI, India, Jan. 23 -- Intellectual Property India has published a patent application (202511122605 A) filed by Chandigarh University, Mohali, Punjab, on Dec. 5, 2025, for 'real-time multi-frame face reconstruction and unique-id system for infiltration detection.'
Inventor(s) include Nitish Sharma; Sachin Chawla; Pratham; Vishal; and Ashi Gupta.
The application for the patent was published on Jan. 23, under issue no. 04/2026.
According to the abstract released by the Intellectual Property India: "The present invention relates to a real-time infiltration detection system that enables accurate identification of individuals from partial or occluded facial views under practical surveillance conditions. The system employs one or more cameras and an edge-processing unit to detect partial facial patches from continuous video streams, associate them across frames, and reconstruct a high-quality facial template using a multi-frame fusion engine. From the reconstructed embedding, the system generates a deterministic, privacy-preserving Unique Face Identifier (UID) through quantization and cryptographic hashing. The UID is securely matched against authorized government identity databases to identify known individuals. When no match is found and reconstruction confidence exceeds a threshold, the system automatically generates and transmits a secure alert packet containing the UID, metadata, and geolocation to the nearest police or security authority. The system further incorporates encrypted communication, tamper-evident audit logging, and optional federated learning to enhance accuracy and privacy in real-time security operations."
Disclaimer: Curated by HT Syndication.
