Extract and Classification of Iris Images by Fractal Dimension and Efficient Color of Iris

last decade, identification by biometric features such as iris and fingerprint has been considered very much. Last introduced methods, in fact, could achieve high accuracy, but one of the most common problems in these methods is the lack of scalability. So these methods are suitable for use in small databases of iris. One solution for this problem is using the hierarchy classification. In this paper, fractal dimension of iris and effective range of color in RGB layers are used as first and second layers of classification in iris images respectively in order to increase the performance of different methods in human identification. The result of simulation on Phoenix database's data shows that this method is suitably efficient in the classification step.