A o el uman Identifier System usin Retina Ima e and u y Clusterin

摘要: What is important for an identification system to be accurate, low cost, fast, and safe. Almost all current methods such as fingerprint, face, hand palm ir recognition, are vulnerable considering plastic surgeries some other changes in finger print, while this not the case retina. Human retina would never change during ones life thus uni ueness of blood vessels pattern most accurate among biometric systems. In paper, we present a new approach human using recognition fuzzy C means clustering algorithm. This method insensitive rotation, rescaling transformation. The Fourier Mellin transform coefficients moments retinal image have been used features extracted our system. To compensate rotational effects scanner, rotation compensator was designed. For optic disc localization, Haar wavelet snakes model used. experimental results show error rate close zero propose method.