Iris Recognition Using Combined Morphological Operations and Hamming Distance Approach

Nowadays technology has salient progress and among this iris recognition attract many attention due to its importance in our life such as security. Even though, many investigation has been done in this field, but it deserves more. So, in this paper a new segmentation method is performed to segment an exact part of eyes. In order to apply this approach, after preprocessing step, at first, local entropy of grayscale image is utilized. Then, rough mask is created in order to segment the textures for the bottom texture and threshold the rescaled image. After that, for smoothing the edges process and closing all the open holes in the object, morphologically close image is employed, and selected a 9-by-9 neighborhood as it was also chosen by local entropy of grayscale image. Finally, for extracting the top and bottom texture, and calculating the texture image, local entropy of grayscale image is utilized, and using Otsu’s method forglobalizing image threshold. Hamming distance measure is applied in order to find similarity degree between two images. We use CASIA-Iris V3 database and our experimental result demonstrate high performance on this database

Keywords:

iris , acquisition , segmentation , biometrics , morphological operators