In the last decade, many recognition and authentication systems based on biometric mesaurements have been proposed. Still
algorithms based on face images are quite appealing for the possibility to easy adapt and taylor a system to many application
domains.
A system for personal identity verification and also recognition is presented. The core engine is a standard correlation-based
matcher performed on iconic representations of face images. Two data sets are used to validate the performances of the whole
system (from data acquisition to recognition): the former is a standard “academic” database (with known acquisition parameters)
similar to the FERET image set, the latter is an “industrial” data set acquired in a real application scenario. Through standard
statistical tests of the recognition results obtained from the two data sets the actual physical limits of the pattern matcher
are clearly shown. Successively also other aspects are taken into account, related to the feature space, allowing to greatly
improve the system performance reaching almost 100% correct recognition.
Several hints for the development of new techniques for identity verification are also suggested.