Extraction of fractal dimension for iris texture

Fractal dimension has not been an effective feature for authentication using the iris. An attempt is made in this paper to explore its effectiveness on iris texture by deriving it from the features obtained using the Information Set based methods. In addition to this, a surface finish measurement tool called Topothesy is also employed for determining the fractal dimension. The well-known Box-Counting method is discussed in brief to facilitate comparison with the new approaches. All the approaches have been tested on iris images from CASIA iris v3 database. The results of comparison with the tested approaches indicate that information set based method and topothesy based method perform better than the traditional methods. The results demonstrate the feasibility of using fractal features to quantify biometric traits like iris feature as well as the superiority of the newly proposed Information set based methods over the traditional approaches.