A pupil center detection algorithms for partially-covered eye image

This study focuses on development of the algorithms that can be applied in video-oculograph method to find a center of a pupil area in captured eye image, where the pupil area is partially covered by eyelashes and eyelids. The eye image captured by monochrome charge-coupled device camera and frame grabber has spatial resolution of 640 pixels by 480 pixels and 8-bit gray levels per pixel. We assumed that the pupil area in the captured eye image was a perfect circle. The center of pupil area could be obtained by applying the proposed algorithms for thresholding of captured gray-scale image, size filtering and noise removing techniques even though the pupil area was partially covered by eyelashes and eyelids.