Possibility theory based approach for iris recognition in biometric systems

Iris recognition could be among reliable means and useful features for advanced biometry technics. In this work, we propose an efficient and novel scheme of one biometric advanced tool using one single modality, i.e the iris, and based on the possibility theory. Such novel tool could easily encompass the variant forms of imperfections that could affect iris data, so reinforcing then the performance of the related recognition system. Accordingly, the use of the possibility theory could be so an attractive tool for modeling such data with its variant forms. Experimental results on identity verification show that the proposed system with possibility theory yields interesting results especially when imperfect biometric data was involved, and have satisfactory performance and robustness in terms of identification.