Mobile ear recognition application
Miloš Oravec, Jarmila Pavlovičová, Dominik Sopiak, Vojtech Jirka, Marek Lóderer, Lubos Lehota, Marek Vodicka, Matej Fackovec, Matej Mihalik, Martin Tomik, Jozef Gerat · 2016
Many technical papers and scientific studies were written about nowadays mobile devices popularity. Also biometrics and especially the face recognition belong already to standard authentication methods. However, the face is not the only one human feature, which contains unique data about person. Shape and folds of a human ear could also be used for reliable identification of people. Information obtained from ear can be used as an extension in multi-modal biometric recognition systems and thereby multiply their effectivity and declarative power. In addition, anybody can own such system implemented on her/his smart device. This paper provides complex comparison of advantages and disadvantages of used approaches for preprocessing, extraction and classification of ear image data. We describe our application flow from the user registration until the user classification. We also show our approach to proposal of an ear biometric application. Finally, we present the results obtained by herein described algorithms and application.