Tactile Fingerprint Sensor Using Piezoresistive Microstructures

Author: PARRAIN F

Advisor: COURTOIS B.

Coadvisor: CHARLOT B.

Thèse de Doctorat INPG, Spécialité Microélectronique

Defense: 02/12/2002

ISBN: 2-913329-98-5

Pages: 162

Abstract:

Nowadays, the need of identifying users is becoming more and more necessary for several typical operations as access controls, workstation login or electronic banking. In this way, many systems as credit cards, mobile phones or computers require keys or alphanumeric passwords. Biometrics recognition is envisaging new solutions using constant features of the user's body with the convenience that they can't be lost, forgotten or stolen. We can give as examples the human speech, the characteristic of the face, the pattern of the iris and so on. Most applications are based on the fingerprint pattern that is the easiest to use. A typical fingerprint covers an area of about 100 mm² and includes several characteristic points so called minutia (generally a number from 12 to 20). Extracting their relative positions, these minutia allow to create a specific signature for each user guaranteeing a secured identification. It is possible to classify fingerprint sensors according to the read mechanism it works (mechanical, optical, capacitance and thermal) and according to the sensor array arrangement (single line, full or partial matrix of pixels). Until now the most used sensing part geometry has been the full or partial array but in this case the sensor is very expensive due to the fact that it covers a large area of silicon. We propose to realize a fingerprint sensor composed by a unique row of microstructures (in fact three in the case of the prototypes presented in this thesis) so as to minimize the size of the chip and consequently its prize. In order to obtain a complete image from this single row of pixels, the user has to pass his finger following a translation movement above the active area of the sensor. In the same time, the different gauges included in the microbeams are scanned. The resistivity change induced by the microstructure deflections (i.e. the finger relief) is then amplified and numerized (8 bit i.e. 256 gray scales) using the integrated electronic interface.

Identifiers: microsystems, simulation