Techn. Fakultät Website deprecated and outdated. Click here for the new site. FAU-Logo

Attila Budai M. Sc.

Alumnus of the Pattern Recognition Lab of the Friedrich-Alexander-Universität Erlangen-Nürnberg

My vision is to develop a fast and robust algorithm to detect glaucoma and other diseases on eye fundus images

Evaluation of Spectral Filtered Fundus Images

The aim of the project is to prove the usefulness of spectral filtered fundus images developed at the eye clinic in Erlangen. These RGB images are similar to the common fundus images, but none of the three channels are under- or oversaturated. Due to the colour balance these images provide more information compared to the common fundus photographs. We have to extract and show this information gain with measurements of features which were not available until now, and we have to adjust the quality measurement algorithms to check the three colour channels one by one, and give a feedback about which filter setting looks optimal for different tasks.

The following pictures show one of the mentioned features: the optical density of macular pigments is calculated from the absorbtion difference in the blue and green visible light domain. The semiautomatic algorithm needs a manually selected ROI which contains the macula region. In this region the algorithm detects the center of the macula, and calculates an optical density profile of the pigments.

Spectral filtered fundus image:
None of the channels are
under- or oversaturated

Normalized absorbtion
difference of blue and
green channels
in the selected ROI

Density profile:
Showing the density changes
as a function of distance
from the macula center