Guest Editorial Special Issue on Mobile Biometrics

Biometrics involves the use of physical and/or behavioural traits for human authentication and identification. It is a well-established and growing discipline supporting many practical applications including cyber security and mobile devices. In recent years, stimulated by hardware advances and a rapidly growing consumer market built around increasingly powerful mobile phones and other portable platforms, mobile biometrics has become an important challenge for biometric applications and will continue to do so. It is predicted that the global market for mobile biometrics will grow substantially in the coming years. Mobile biometrics aims to achieve the functionality and robustness of conventional biometrics while also supporting portability and mobility in order to bring greater convenience, flexibility, and opportunity for its deployment in a wide range of operational environments ranging from consumer applications to law enforcement. However, achieving these aims brings new challenges, such as issues about power consumption, algorithmic complexity, device memory limitations, frequent changes in operational environment, security, durability, reliability, connectivity, and so on. This Special Issue (SI) aims to bring together some good examples of current research that addresses some of the major challenges for mobile biometrics. The SI provides a platform for both academic researchers and industry partners to become familiar with the latest new research, stimulating discussion on existing and emerging challenges in mobile biometrics, and raising awareness of potential solutions to advance research in mobile biometrics. After careful review and selection, this SI has accepted four papers from those submitted. These four papers cover mobile biometrics from different aspects, which is detailed in the following. The first paper, ‘A Method for Using Visible Ocular Vasculature for Mobile Biometrics’, by V. Gottemukkula, S. Saripalle, S. Tankasala, and R. Derakhshani, investigates the use of existing visible light cameras in mobile phones for vascular pattern segmentation, extraction and recognition. Several stages are involved in the biometric process, including interest point detection, feature description, and matching, with Gabor phase filters used for feature extraction and matching. Low error rates are reported in the experimental validation of the approach. The study reported shows the feasibility and utility of extracting vascular patterns for user recognition in mobile devices. The second paper, ‘An Authentic Mobile-Biometric Signature Verification System’, by F. Zareen and S. Jabin, presents an overview of typical mobile biometric systems, including different devices and hardware to support mobile biometrics. The paper discusses open issues and challenges for developing mobile biometric systems. To that end the paper describes a mobile biometric system for signature verification, which is supported by experimental validation. The third paper, ‘The Blind Subjects Face Database (BSFD)’, by N. Poh, R. Blanco-Gonzalo, R. Wong and R. Sanchez-Reillo, introduces a new face database captured from blind subjects. The underlying application scenario is to use the face characteristic to unlock a mobile device, which is convenient and fast for real applications. However, one needs to address how well this technique can be used for visually impaired people. The collected database, BSFD, facilitates the usability and face recognition studies involving people with varying degrees of visual impairment for mobile applications. More importantly, the database is publicly available for use by other researchers to investigate related issues further. The last paper, ‘Small Fingerprint Scanners Used in Mobile Devices, The Impact on Biometric Performance’, by B. Fernandez-Saavedra, R. Sanchez-Reillo, R. Ros-Gomez, and J. Liu-Jimenez, examines the performance of fingerprint recognition in mobile devices where the embedded fingerprint scanners could have different sensing areas, capturing fingerprint images of different image quality. Both public and commercial algorithms (including a database specifically assembled for this study) are used for the testing and performance evaluation. Experimental results show that the fingerprint matching performance is affected by the scanner sizes, which are correlated to the fingerprint image quality.