Once confined to indoor biometric applications depending on dedicated acquisition devices, recently the iris has proved to be a suitable biometric for in-the-wild ubiquitous person authentication, thanks to continuously improving image capturing/processing performances provided by last generations of smartphones. In this mobile context, the efficiency of the whole processing pipeline represents a crucial aspect of any practical application and the segmentation task, that is deeply affected by noisy iris images may become a serious bottleneck. This work presents F-FID, an effective and time-wise efficient approach to de-noising of iris images by means of a fuzzy controller without sacrificing their resolution and saliency. The experiments, specifically conducted on the MICHE dataset, confirm that the proposed method provides segmentation accuracy comparable to that achieved by state of the art algorithms, while requiring less than twenty percent of their average computing time.
F-FID: fast fuzzy-based iris de-noising for mobile security applications
Barra, Silvio
;
2019-01-01
Abstract
Once confined to indoor biometric applications depending on dedicated acquisition devices, recently the iris has proved to be a suitable biometric for in-the-wild ubiquitous person authentication, thanks to continuously improving image capturing/processing performances provided by last generations of smartphones. In this mobile context, the efficiency of the whole processing pipeline represents a crucial aspect of any practical application and the segmentation task, that is deeply affected by noisy iris images may become a serious bottleneck. This work presents F-FID, an effective and time-wise efficient approach to de-noising of iris images by means of a fuzzy controller without sacrificing their resolution and saliency. The experiments, specifically conducted on the MICHE dataset, confirm that the proposed method provides segmentation accuracy comparable to that achieved by state of the art algorithms, while requiring less than twenty percent of their average computing time.File | Dimensione | Formato | |
---|---|---|---|
Barra2019_Article_F-FIDFastFuzzy-basedIrisDe-noi.pdf
Solo gestori archivio
Tipologia:
versione editoriale (VoR)
Dimensione
2.91 MB
Formato
Adobe PDF
|
2.91 MB | Adobe PDF | Visualizza/Apri Richiedi una copia |
I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.