Fingerprint recognition for biometric passwords

The goal of this research is the design and implementation of an integrated system for personal identification which uses fingerprint recognition. The system is made of a sensor, a classification and a validation systems. Some important issues to be addressed are: compactness, low-cost and reliability. For the above reasons we have faced the problem by studying and implementing a non-optical fingerprint imager which uses a capacitive sensor cell array. A 390 dpi (dot per inch) prototype has been successfully implemented in a general purpose CMOS technology. The sensor senses the distance of the skin surface by means of capacitive sensing made by cell arrays arranged in $200$ rows by $200$ columns. The primary goal is to get two dimensional images whose gray level function is linear with respect to the distance of the skin surface to the sensor array. To achieve this goal, several circuit techniques have been proposed, simulated and implemented. The measured result has confirmed the models and several further applications can be found for this kind of sensor: from pressure to acceleration sensors from micro-machinery to integrated microphones. The classification system enhances the sensor image and identifies the fingerprint image, comparing it to a previously stored one. The result of the classification consists in accepting the two images as those belonging to the same person, or rejecting such a statement. The fingerprint identification is based on the comparison of feature points in both the images and on the matching of the regions surrounding these areas. The feature points are the standard ending/bifurcation points of the skin ridges.

Researchers:

Microelectronics Group Home Page Staff
Research Activities Teaching Activity
Publications Local Stuff
DEIS Home Page University of Bologna Home Page

External link, Local link, For local sites only, Stiil empty.