Optical sensor measurement and biometric-based fractal pattern classifier for fingerprint recognition

This paper proposes biometric-based fractal pattern classifier for fingerprint recognition using grey relational analysis (GRA). Fingerprint patterns have arch, loop, whorl, and accidental morphologys, and embed singular points, which result in establishing fingerprint individuality. An automatic fingerprint identification system consists of three stages: image acquisition and processing, feature extraction, and pattern recognition. Fingerprint images are captured from subjects using an optical fingerprint reader (OFR). Digital image preprocessing (DIP) is used to refine out noise, enhance the image, convert to binary image, and locate the reference point. For binary images, Katz’s algorithm is employed to estimate the fractal dimension (FD) from two-dimension (2D) image. Biometric characteristics are extracted as fractal patterns using Weierstrass cosine function (WCF) with different FDs. GRA performs to compare the fractal patterns among the small-scale database. For 30 subjects in the laboratory, the proposed classifier demonstrates greater efficiency and higher accuracy in fingerprint recognition.