Electronic documents (passport, electronic identity card CIE 3.0) are equipped with a microprocessor which stores biometric data (face image and fingerprints) of the document holder, in addition to other personal data. Such biometric samples can be used for automatic identity verification, i.e. to prove the correspondence between the document owner and the person who exhibits it. Such correspondence is assessed by comparing the biometric data stored in the chip against the biometric data acquired live (biometric data of the person presenting the document). Even in this controlled scenario, the performance of the identity verification system is related to the quality of the biometric data stored into the document. The importance of the quality factor is confirmed by the provisions of ISO/IEC 39794-5, which defines the possibility of storing into the document a biometric sample quality value in dedicated quality blocks.
The research activity in conjunction with Istituto Poligrafico Zecca dello Stato is focused on the design of a solution for automatic face image quality assessment to be applied to electronic ID documents issued by the Italian State. Face image quality estimation is still an open problem since, unlike what happens for other biometric characteristics such as fingerprints, no standard definitions are available yet. The problem is even harder when the objective is quality assessment in the context of electronic ID documents where images are fully ISO compliant and characterized by an overall very high quality; for this kind of images the general quality indicators available in the literature tend to provide a flat score that doesn’t provide significant information. This research activity is aimed at defining a quality assessment technique for high-quality images, able to predict the utility of a specific image sample for face verification purposes.
A further research topic addressed within this collaboration is related to face morphing.