Detection of abnormalities in retinal images

Abstract

The human eye is the organ which gives us the sense of sight. The eye reflects or emits the light to interpret the shapes, colors, and dimensions of objects in the world. Retina places a major role in the human vision system. The retina gets affected by long-term dietetic mellitus called Diabetic Retinopathy (Microaneurysm). Earlier detection of these abnormalities will prevent the vision loss. This work aims to detect such abnormalities in the retinal image and to classify them based on their severity. To detect the abnormality, two preprocessing and one candidate extraction method are proposed and various stages of abnormalities are classified based on the features like area, mean, standard deviation, entropy etc. Adaptive Neuro Fuzzy Inference System (ANFIS) is an effective tool used for effective screening of retinal abnormalities. ANFIS is used to classify the retinal images as normal, mild, severe depending on their severity.