Specular Reflection Removal Using Morphological Filtering for Accurate Iris Recognition

Specular reflection is the brightest point existing in the iris image. The light source used for iris image acquirement and reflective surface of the cornea are the major causes of specular reflection. Specular reflection affects the iris segmentation module of iris recognition systems and increases the non-match rate. Hence, there is a need for removing specular reflections in iris images. In this paper, a specular reflection removal approach for accurate iris segmentation is proposed. This method was executed and evaluated using iris images collected from the UBIRIS.v1. database. The experimental outcomes clearly show that the average Structural Similarity Index (SSIM) between the output reflection removed iris image and the input iris image is 0.9962.