An image processing method for improved pupil size estimation accuracy

Pupil size is adjusted through the contraction of the iris muscles that are controlled by the autonomic nervous system. The oscillation of the pupil size and the profile of the pupillary light reflex are important parameters that can reflect certain conditions of the autonomic system such as cognitive assignment, emergency, and autonomic dysfunctions. This study sought to develop a device that enables automated and accurate estimation of the papillary size, through which the pupil size variability (PSV) and the profile of the papillary light reflex are induced. Methods for blink image detection that needs preprocessing, pupil size estimation in a blink image, pupil segmentation in a various conditions, and transformation of coordinates to return a distorted pupil projected on a two-dimensional (2D) image plane to its original size were presented. In addition, the image acquisition system and analysis of factors affecting the estimation of pupil size were discussed.