NOISY FINGERPRINTS CLASSIFICATION WITH DIRECTIONAL FFT BASED FEATURESUSING MLP

A methodology is described for classifying noisy fingerprints directly from raw unprocessed images. The directional properties of fingerprin ts are exploited as input features by computing one-dimensional fast F ourier transform (FFT) of the images over some selected bands in four and eight directions. The ability of the multilayer perceptron (MLP) f or generating complex boundaries is utilised for the purpose of classi fication, The superiority of the method over some existing ones is est ablished for fingerprints corrupted with various types of distortions, especially random noise.