Information Technologies and Computer Engineering

The article analyses the known methods of dynamic signature verification, which are summarised in a classification table. A method of dynamic signature verification based on a spiking neural network is proposed. Three dynamic signature parameters l(t), Dα(t), Z(t) are selected, which are invariant to the signature angle, and after their normalisation - also to the spatial and temporal scales of the signature. These dynamic signature parameters are fed to the spiking neural network for recognition simultaneously in the form of time series without prior conversion to a vector of static features, which, on the one hand, simplifies the method due to the absence of complex computational conversion procedures, and on the other hand, prevents the loss of useful information, and therefore increases the accuracy and reliability of signature verification and recognition (especially when recognising forged signatures that are highly correlated with the originals). The neural network used has a simple training procedure, and not all neurons of the network are trained, but only the initial ones. If you need to add new signatures, you don't need to retrain the entire network, but rather add a few output neurons and train only their connections

References