Diabetic retinopathy lesion detection using region-based approach

Diabetic Retinopathy (DR) is one of the disease due to complication of diabetes. Late in treatment of this disease can cause blindness. Therefore early eye screening to detect the existing of the abnormality in retina is very crucial for further treatment. There are several abnormalities or lesions as sign of DR disease such as microaneurysm, hemorrhage, soft exudates and hard exudates. In this article automated method using region based approach was proposed to identify abnormal region on fundus image. The approach involves region feature extraction using haralick texture features and two different classifier; support vector machine (SVM) and multi layer perceptron (MLP) with different experiment. The advantageous of this approach are not involving normal structure segmentation process as standard procedure applied by many researchers. The proposed approach was tested on several types dataset and the result show promising result.