Volume1,May 2016,

Topic : Detection of Hard Exudates, Soft Exudates, Red Spots and Hemorrhages

Authors:S.Saranya devi || P.Anuja,Bhagyalakshmi.K.M, Jinchu.J

Abstract:The most complex system in human body is eye. Nowadays there is increase in vision loss patients mainly due to retinal diseases. Some diseases cannot be identified early unless it affects the vision as it has no symptoms. But it can be identified by some clinical signs in the retina which cannot be viewed through naked eye but through fundus photography. Traditionally, retinal diseases are diagnosed by manual observations of fundus images and it is a time consuming process. So an automatic algorithm is need to be develop to detect and classify various clinical signs. Some of the clinical signs of retinal diseases are hard exudates, soft exudates, red spots and haemorrhage. We used MATLAB software to develop an algorithm. Various techniques have been proposed to detect these signs but the efficiency is less and there is a chance of mismatching. Hence we proposed a new technique known as skin locus model which is used only for face detection till now. In this technique pixels between the images are compared and based on certain threshold the various clinical signs are detected ad classified. We used DRIVE database, an open source database images, which has been acquired from the fundus photography. Our proposed algorithm reduces man power, time consumption and also help the retinal diseased patients by preventing or delaying the vision loss. It also avoid mismatching of clinical signs..

Keywords: CLINICAL SIGNS AND DETECTION METHODS

Download Paper