MUMBAI, India, Aug. 17 -- Intellectual Property India has published a patent application (202611081021 A) filed by Ajay Kumar Garg Engineering College on July 01, 2026, for System For Rotation-Invariant Fingerprint Recognition And Statistical Template Generation And Working Method Thereof.
Inventors include Dr. Akhilesh Verma; and Dr. Parul Verma.
The application for the patent was published on August 14, 2026, under issue no. 33/2026.
Abstract: The present invention discloses a system for rotation-invariant fingerprint recognition and statistical template generation and working method thereof. The system comprises a biometric acquisition module (100) configured to capture a fingerprint image, an image pre-processing unit (110) configured to enhance ridge structures, a minutiae extraction and post-processing engine (120) configured to identify and filter fingerprint minutiae, and a bounding-box normaliser (130) configured to normalize minutiae coordinates onto a predefined coordinate space. A centroid computation unit (140) computes a geometric centroid for generating centroid-aligned multi-resolution grids through a multi-resolution grid generator (150). A rotational sweep and accumulation engine (160) performs rotational transformation and accumulation of minutiae distributions across multiple angular positions. A statistical moment extractor (170) derives variance, skewness, and kurtosis values, and a feature vector concatenator (180) generates a fixed-length fingerprint template. A similarity scoring and decision unit (200) performs fingerprint authentication or identification. The invention provides compact, training-free, and inherently rotation-invariant fingerprint recognition without explicit alignment. Accompanied Drawings [Fig. 1-4]
Disclaimer: Curated by HT Syndication.
