Abstract

Iridology is the science and practice that helps to learn the health status through the examination of the structural aspects of an eye. Iris segmentation is an important phase in Iridology that aids to recognize iris signs and identify aberrant organ activity. The primary elements of iris segmentation are the detections of the iris and pupil. However, iris segmentation is greatly impacted by noises and images with low quality. The proper method for iris and pupil detection supports improving the result of iris segmentation in Iridology. This study provides a computer vision-based method to segment the human eye’s iris and pupil using an alternate method. The procedure is based on the red channel information of the image with various image processing operations. Iris images from the UBIRIS.v1. Dataset, which was taken in less restrictive image acquisition conditions, is used in an experiment. This method gives a unique way for iris segmentation with an accuracy of 97.49%.

The International Conference on Artificial Intelligence and Computer Vision

AICV 2023: The 3rd International Conference on Artificial Intelligence and Computer Vision (AICV2023), March 5–7, 2023 pp 231–241