Fundus image diseases classification using convolutional neural network

Abstract

Nowadays, the number of people affected by the various types of Eye disease keeps on increasing rapidly. The highest stage a person can get affected by an eye disease is blindness. The fact is that we cannot be able to stop this, but we can control and gradually decrease the percentage of people affected by eye disease. This can be done by detecting what type of eye disease the person is affected earlier and taking necessary treatment accordingly. For that, we developed software that can identify three types of eye disease using fundus image. So the classification of eye disease can be done by much technology. Here we are using a convolution neural network for classification. The image processing suits well for segmenting and analyzing the image according to the user’s requirement. The outcome may be best described by identifying disease types and stages. In existing, different machine learning techniques were used to detect only a particular disease. Here we are using CNN, which is a part of deep learning systems, and so that we are going to classify two more additional diseases.