APPLICATION OF THE MORPHOLOGICAL ULTIMATE OPENING TO THE DETECTION OF MICROANEURYSMS ON EYE FUNDUS IMAGES FROM A CLINICAL DATABASE

摘要

Diabetic Retinopathy (DR) is a severe disease which can cause blindness.OPHDIAT is a telemedicine network for DR mass screening,which has gathered thousands of clinical high resolution color eye fundus images.The TELEOPHTA project has been launched in order to develop a computer aided diagnosis system of DR,which aims at performing a preliminary analysis of the OPHDIAT images in order to filter most images corresponding to healthy eyes.Microaneurysms (MAs) are likely to be the lesions present at the earliest stage of the disease.In this paper,a new method of MAs detection,using the recently proposed ultimate opening,is presented.The proposed method does not use any supervised classification,while provides a competitive and efficient way to detect MAs,especially for our clinical database.Further improvements may be brought through the accurate detection of the retinal elements and other retinal diseases,or through the estimation of the image quality.