I’m embarking on a project using pupillometry (the study the pupil within the eye).<p>My requirement is to use computer vision to determine the size of an a closeup image of a pupil (for accuracy purposes the pupil needs to take up 80% of the image).<p>I can’t use a reference object in the image.<p>I could use the LIDAR in the new iPhone (unreleased) but the people using this product will not all have the latest iPhone so a more mainstream solution is required.<p>Maybe there is a published scale standard that I’m missing? Maybe there’s an ML dataset of eyes and measurements that I’m missing? Maybe I’m missing some other obvious solution?<p>Any help/input would be greatly appreciated :)
some resources you may want to use:<p><a href="https://news.cornell.edu/stories/2018/10/using-smartphone-cameras-track-alertness" rel="nofollow">https://news.cornell.edu/stories/2018/10/using-smartphone-ca...</a><p><a href="https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4950215/" rel="nofollow">https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4950215/</a><p><a href="https://www.ncbi.nlm.nih.gov/pmc/articles/PMC6166694/" rel="nofollow">https://www.ncbi.nlm.nih.gov/pmc/articles/PMC6166694/</a><p><a href="https://www.youtube.com/watch?v=AVQzu9AL3IU" rel="nofollow">https://www.youtube.com/watch?v=AVQzu9AL3IU</a><p><a href="https://www.researchgate.net/post/What_is_the_best_measure_to_track_pupil_dilation_in_an_experiment_where_a_video_continuously_recorded_one_of_the_participants_eye" rel="nofollow">https://www.researchgate.net/post/What_is_the_best_measure_t...</a>