The Research And Realization Of The Face And Iris Biometric Identification Authentication System

Multimodal (such as face and iris, etc.) biometric identification technology is the field of pattern recognition in recent years, a very hot topic, which people own a variety of comprehensive utilization of biometrics, for today and provide more secure authentication methods . Researchers have been more and more attention.In this paper, the existing single-mode biometric system deficiencies, face and iris recognition on the composition of the multi-modal biometrics to do in-depth study, the main research work is as follows:1,The PCA-based and Adaboost face recognition algorithm were in-depth study;2, The Adaboost and SVM-based iris recognition algorithm were in-depth study;3, The various face-iris recognition fusion algorithms models were more in-depth study; According to actual needs, propose a new hybrid fusion model. Feature level fusion based on the actual algorithm is given.4, designed based on the DM642 + ARM2440, dual-core structure, embedded, face - recognition, integration, and multi-functional, the prototype system. Gives, in the development process, the resulting system, the needs analysis; hardware platform design, circuit and circuit board design; and software design flow, and UML modeling diagram.Also demonstrated in the development process, designed to verify the algorithm, PC-based human face - iris recognition algorithm verification experiment system.