Development on this project had been abandoned but over the last week or so I've started working on it again and already made significant progress.
Finding the pupil is automated now - although I have plenty of ideas for increasing its accuracy. Finding the iris boundary shouldn't be too far behind, currently you have to specify a radius for it in the code.
The 'rubber mapping' to unwrap the iris has a workable version. Really all that's needed after the iris boundary is the image import - I have a camera capable of taking images in the IR spectrum although Linux support for it seems lacking. Then we'll have a workable tool with minimal work on the GUI.... read more