MUMBAI, India, June 22 -- Intellectual Property India has published a patent application (202641045901 A) filed by Saveetha Engineering College on April 10, 2026, for A Real-Time Face Recognition System With Enhanced Accuracy.

Inventor includes Dr. S. Sasi Kumar.

The application for the patent was published on June 12, 2026, under issue no. 24/2026.

Abstract: The present invention relates to a real-time face recognition approach designed to achieve enhanced accuracy under unconstrained and dynamic conditions. The invention addresses the limitations of conventional face recognition techniques, which often suffer from reduced reliability due to variations in illumination, pose, facial expressions, occlusions, and background complexity. By integrating advanced image preprocessing, robust face detection and alignment, and deep learning-based feature extraction, the proposed approach enables accurate and efficient facial identification from live video streams. The invention operates by capturing facial images in real time and applying adaptive preprocessing techniques to normalize lighting conditions and reduce noise. Detected faces are aligned using facial landmarks to minimize pose variations and ensure consistency in feature representation. Discriminative facial features are then extracted using optimized deep learning models, producing compact and distinctive embeddings that accurately represent individual identities. These embeddings are compared against stored facial templates using similarity-based matching to generate reliable recognition decisions with minimal latency. The approach is scalable and adaptable, allowing deployment across edge devices, embedded systems, and cloud-based platforms. Applications include security and surveillance, access control, automated attendance systems, smart environments, and human-computer interaction. By achieving a balance between speed, accuracy, and robustness, the invention advances the state of the art in real-time biometric face recognition technology.

Disclaimer: Curated by HT Syndication.