Neural network verification of dynamic signatures using pressure sensitive pen

In this paper, we present a technique to classify signatures produced by the pressure exerted on the pen tip. Before the features are exerted, a low pass filter using Sum Filter is designed to remove frequencies greater that 50 Hz. A new segmentation technique is used to divide the time series data into segments. The autoregressive (AR) coefficients are derived from each segment. From the coefficients, the power spectral density (PSD) is determined for every segment. Values from the spectral are then fed into a Multilayer Perceptron (MLP) classifier with one hidden layer for verification. A database of 1000 signatures is used for training and testing. The system is tested for genuine as well as forged signatures. The result obtained showed 2.13% error in rejecting genuine signatures and 3.40% error in accepting forged signatures.