EAI Endorsed Transactions on Ambient Systems 17(13): e2

Research Article

Access Control in Smart Homes by Android-Based Liveness Detection

Download53 downloads
  • @ARTICLE{10.4108/eai.17-5-2017.152546,
        author={Susanna Spinsante and Laura Montanini and Veronica Bartolucci and Manola Ricciuti and Danny Pigini and Ennio Gambi},
        title={Access Control in Smart Homes by Android-Based Liveness Detection},
        journal={EAI Endorsed Transactions on Ambient Systems},
        volume={17},
        number={13},
        publisher={EAI},
        journal_a={AMSYS},
        year={2017},
        month={5},
        keywords={liveness detection, spoofing, face recognition, Android, stereo vision},
        doi={10.4108/eai.17-5-2017.152546}
    }
    
  • Susanna Spinsante
    Laura Montanini
    Veronica Bartolucci
    Manola Ricciuti
    Danny Pigini
    Ennio Gambi
    Year: 2017
    Access Control in Smart Homes by Android-Based Liveness Detection
    AMSYS
    EAI
    DOI: 10.4108/eai.17-5-2017.152546
Susanna Spinsante1,*, Laura Montanini1, Veronica Bartolucci1, Manola Ricciuti1, Danny Pigini1, Ennio Gambi1
  • 1: Dipartimento di Ingegneria dell’Informazione, Universita’ Politecnica delle Marche, Via Brecce Bianche 12 Ancona, 60131, ITALY
*Contact email: s.spinsante@univpm.it

Abstract

Technologies for personal safety and security play an increasing role in modern life, and are among the most valuable features expected to be supported by so-called smart homes. This paper presents a low-complexity Android application designed for both mobile and embedded devices, that exploits the available on-board camera to easily capture two images of a subject, and processes them to discriminate a true 3D and live face, from a fake or printed 2D one. The liveness detection based on such a discrimination provides anti-spoofing capabilities to secure access control based on face recognition. The limited computational complexity of the developed application makes it suitable for practical implementation in video-entry phones based on embedded Android platforms. The results obtained are satisfactory even in di erent ambient light conditions, and further improvements are being developed to deal with low precision image acquisition.