Automated microaneurysm detection using mathematical morphology

Diabetes is one of the most rapidly increasing health threats worldwide. One of the further abnormalities is on retina (diabetic retinopathy). Early treatment can be conduct from detection of microaneurysms. The main concentration of this paper is the algorithm to detect microaneurysm with mathematical morphology. The mathematical morphology is choosen because microaneurysms tend to have typical shape. Generally, the algorithm is consist of three stages. The first is preprocessing, the second is detecting candidate microaneurysms, and the third is postprocessing handling the process of removing unused features. The performances is evaluated using the database from DIARETDB1 which provides ground truth collected from several experts and a strict evaluation protocol. The optimal performance will be satified when considering green channel obtaining, PAL size image processing, adaptive histogram equalization threshold at 0.03, canny edge detection threshold at 0.16, MAs and optimum microaneurysms size at 5 to 16 pixels.