A neural network approach is proposed to build the first stage of an A
utomatic Handwritten Signature Verification System. The directional Pr
obability Density Function was used as a global shape factor and its d
iscriminating power was enhanced by reducing its cardinality via filte
ring. Various experimental protocols were used to implement the backpr
opagation network (BPN) classifier. A comparison, on the same database
and with the same decision rule, shows that the BPN classifier is cle
arly better than the threshold classifier and compares favourably with
the k-Nearest-Neighbour classifier.