Design of a Multi-biometric Platform, based on physical traits and physiological measures: Face, Iris, Ear, ECG and EEG

Security and safety is one the main concerns both for governments and for private companies in the last years so raising growing interests and investments in the area of biometric recognition and video surveillance, especially after the sad happenings of September 2001. Outlays assessments of the U.S. government for the years 2001-2005 estimate that the homeland security spending climbed from $56.0 billions of dollars in 2001 to almost $100 billion of 2005. In this lapse of time, new pattern recognition techniques have been developed and, even more important, new biometric traits have been investigated and refined; besides the well-known physical and behavioral characteristics, also physiological measures have been studied, so providing more features to enhance discrimination capabilities of individuals. This dissertation proposes the design of a multimodal biometric platform, FAIRY, based on the following biometric traits: ear, face, iris EEG and ECG signals. In the thesis the modular architecture of the platform has been presented, together with the results obtained for the solution to the recognition problems related to the different biometrics and their possible fusion. Finally, an analysis of the pattern recognition issues concerning the area of videosurveillance has been discussed.