A new detection method of singular points of fingerprints based on neural network

Abstract

Singular points detection, a crucial step for fingerprint identification system, is accurately robust, and reliable. In the processing of the fingerprint image matching and classification, many method use singular points to align two fingerprint images to surmount the problems about rotation and translation. The performances of the fingerprint recognition system rely on the effect of singular points detection. This paper introduces a new method through backpropagation neural network (BPNN) to detect singular points of fingerprint images. The algorithms that directly make use of the gray-scale of fingerprint images to detect singularities are different from the other methods, which are based on the neural network. The algorithm is tested accurately and reliably for a great many of fingerprint images in the FVC2002.