In this work, algorithms for extracting global geometric and local gri
d features of signature images were developed. These features were com
bined to build a multi-scale verification function. This multi-scale v
erification function was evaluated using statistical procedures. Resul
ts indicated that the multi-scale verification function yielded a lowe
r verification error rate and higher reliability than the single-scale
verification function using either global geometric or local grid fea
ture representation. The correct verification rate of the multi-scale
system was more than 90% in rejecting skilled forgeries and was perfec
t in rejecting simple forgeries based on a limited database.