Face image quality estimation is still an open issue since, unlike what happens for other biometric characteristics such as fingerprints, no standard definitions are available yet. The problem is even harder when the focus of quality assessment is the context of electronic ID documents for which, according to the provisions of ISO/IEC 39794-5, a quality value will be stored in the future in dedicated quality blocks. In case of high-quality images, the general indicators available in the literature tend to assign a flat score that does not contribute to provide significant information. This work documents a study aimed at defining a quality score indicator for high-quality images, able to predict the utility of a specific image for face verification purposes. A quality regressor is proposed, based on a large set of quality elements including ISO/ICAO controls and quality scores provided by deep-learning based solutions. A number of experiments highlight specific issues to be addressed in this scenario and confirm the effectiveness of the proposed approach with different face recognition systems.

Franco A., Magnani A., Maltoni D., Maio D., Odorisio L., De Maria A. (2022). Face Image Quality Assessment in Electronic ID Documents. IEEE ACCESS, 10, 77744-77758 [10.1109/ACCESS.2022.3191463].

Face Image Quality Assessment in Electronic ID Documents

Franco A.
Primo
Conceptualization
;
Maltoni D.
Conceptualization
;
Maio D.
Supervision
;
2022

Abstract

Face image quality estimation is still an open issue since, unlike what happens for other biometric characteristics such as fingerprints, no standard definitions are available yet. The problem is even harder when the focus of quality assessment is the context of electronic ID documents for which, according to the provisions of ISO/IEC 39794-5, a quality value will be stored in the future in dedicated quality blocks. In case of high-quality images, the general indicators available in the literature tend to assign a flat score that does not contribute to provide significant information. This work documents a study aimed at defining a quality score indicator for high-quality images, able to predict the utility of a specific image for face verification purposes. A quality regressor is proposed, based on a large set of quality elements including ISO/ICAO controls and quality scores provided by deep-learning based solutions. A number of experiments highlight specific issues to be addressed in this scenario and confirm the effectiveness of the proposed approach with different face recognition systems.
2022
Franco A., Magnani A., Maltoni D., Maio D., Odorisio L., De Maria A. (2022). Face Image Quality Assessment in Electronic ID Documents. IEEE ACCESS, 10, 77744-77758 [10.1109/ACCESS.2022.3191463].
Franco A.; Magnani A.; Maltoni D.; Maio D.; Odorisio L.; De Maria A.
File in questo prodotto:
File Dimensione Formato  
Face_Image_Quality_Assessment_in_Electronic_ID_Documents.pdf

accesso aperto

Tipo: Versione (PDF) editoriale
Licenza: Licenza per Accesso Aperto. Creative Commons Attribuzione (CCBY)
Dimensione 1.27 MB
Formato Adobe PDF
1.27 MB Adobe PDF Visualizza/Apri

I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.

Utilizza questo identificativo per citare o creare un link a questo documento: https://hdl.handle.net/11585/902433
Citazioni
  • ???jsp.display-item.citation.pmc??? ND
  • Scopus 1
  • ???jsp.display-item.citation.isi??? 0
social impact