Here we show an application of an identity recognition system based on face biometrics using the EBGM technique. The system works in two distinct modalities: enroll and recognition.

During the enroll stage the tool acquires, process' and stores a set of biometric data belonging to the user. The system enrolls a user with just one photo of  his/her face, this is done with a processing phase during which a set of distinctive features is extracted. These features are formatted into a data structure (called template)  which is stored in a database and used in the next stages.

During the recognition stage, the user has just to let the system acquire an image containing the face. The tool now can extract a new set of distinctive features and therefore can create a new data structure (called pattern) and search the database to find (if any) a similar user. Here the concept of similarity is represented by  a score associated to the comparison between the pattern and the stored templates. Comparing the scores the software can select the best candidates as the registered user for whom the computed score is maximum. To finally decide for recognition the best candidate has to have a score above a decision threshold.