Abstract:
Diabetic retinopathy (DR) is a serious issue found among people with diabetes which
cause severe damage to retina of eye and may lead to complete blindness.
Diagnosis of diabetic retinopathy in early stages may be treated to avoid complete
blindness. The objective ofthis project is to develop a system which will classify
images into five stages of retinopathy by recognizing image features. The algorithm
we are using in it is SVM. The main advantage of this technique is that it provides
extraction and detection of features that are suitable for disease recognition in eye
images. The system will be able to serve the automatic detection of blindness due to
diabetes at the early stages.
