Menú Accesibilidad

In-air signature, new authentication technique for mobile phones

A system created by Group of Biometrics, Biosignals, and Security (GB2S) from CeDInt at UPM allows verifying the identity of mobile users through their in-air signature.

This new biometric authentication technique provides higher security than the use of a PIN (which can be guessed, lost, copied or stolen).  It can be applied to the e-commerce field, protecting transactions which require the verification of the user identity or preserving private information stored in the mobile phone. that should not be accessible from anyone except the user.

The spread of a new generation of mobile phones and the drop of prices in the internet mobile connections, provide us the possibility to reach a great number of services  everywhere. Actually, some of the world's leading companies propose nowadays the use of the phone as a new method of payment (“Google Wallet”).

One of the research lines of GB2S is to increase the security in these services and transactions by using biometrics and/or crptographic techniques.

In this case we propose an innovative biometric technique based on the in-air signature made with a mobile phone embedding an accelerometer. From this sensor the biometric system is able to extract the accelerations of the signature in the different axis. This information is used to verify the identity of the users and allow them to carry out the protected transaction.

The researchers of GB2S have conducted several experiments in order to assess the accuracy of in-air signature technique. Concretely, they have created a database with more than 100 samples of in-air signatures from different individuals of UPM. These subjects made their usual handwritten signature in the air holding a mobile device while they were recorded with a videocamera. Afterwards, six people analysed these recordings and attempted to imítate each signature in order to try to forge them.

Using this database, the researchers have proposed and evaluated different algorithms to maximize the accuracy of the system and to minimize the number of forged signatures accepted,  reaching an error rate around 3%. This result shows the viability of in-air signature as a biometric technique, especially taking into account that forgers were able to view the recordings several times.

The main aim of GB2S is to provide methods that make it possible to increase the security of common people. For this reason, this group is working in other biometric techniques like hand shape, iris, fingerprint, face, gait and scent. The fusion of these techniques with cryptography has also derived new cryptobiometric techniques which allow generating new cryptographic keys from biometric features. Furthermore, GB2S is also working in several projects dealing with biosignals. For instance they are able to detect stress from physiological signals and to identify different diseases like cancer from the scent of the breath.

G. Bailador del Pozo, C. Sánchez Ávila, J. Guerra Casanova & A. de Santos Sierra, “Analysis of pattern recognition techniques for in-air signature biometrics”, Pattern Recognition, 44 (10-11), pp.: 2468-2478, 2011.

J. Guerra Casanova, C. Sánchez Ávila, A. de Santos Sierra & G. Bailador del Pozo, “Score optimization and template updating in a biometric technique for authentication in mobiles based on gestures”, Journal of Systems and Software, 84 (11), pp.: 2013-2021, 2011.