Localization of Singular Points in Fingerprint Images Based on the Gaussian-Hermite Moments
基于Gaussian-Hermite矩的指纹奇异点定位

In most fingerprint classification and identification algorithms, extracting the number and precise location of singular points (core and delta) is of great importance. In this paper, an adaptive algorithm for singular points detection is proposed, which is based on the behavior of Gaussian-Hermite moments. In order to detect singular point accurately, the distribution of Gaussian-Hermite moments of different orders of the fingerprint image in multiple scales is used. A PCA-based (principal component analysis) method is used to analyze the distribution of Gaussian-Hermite of fingerprint image. Experimental results show that the proposed algorithm is able to locate singular points in fingerprint with high accuracy.