BIOMETRIC KEYSTROKE RECOGNITION BASED ON HYBRID SVD AND WAVELET FOR FEATURE TRANSFORMATION

Abstract

The main aim of this work is to use a keystroke biometric system as a behavioral type of biometrics and to improve the accuracy and dependability of the system. In this proposed we've pre-processed the data of dynamic keystrokes by converting the feature to a one-dimensional vector. In feature extraction, we've used Wavelet Energy (WE) by implementing 2D dimensional Discrete Wavelet (2D-DWT) into four-level and computing the energy for the Singular Value Decomposition (SVD). SVD is computed on the result of the wavelet and saved in a file for training information. Wavelet transforms Daubchies “DBI” basic function has the advantage that provides a good energy localization in the frequency domain as other wavelet transforms and then using Elman networks (Backpropagation) for training and testing the system and it is useful in such areas as signal processing and prediction where time plays a dominant role.