Published Paper Details:

The main factor in senior blindness is diabetic retinopathy, a disorder of the retina that affects those with diabetes. It is an asymptomatic condition marked by blood vessel irregularities that could produce in bleeding or fluid leakage and distort vision. Therefore, blood vessel removal is essential in helping ophthalmologists identify this condition early and avoid vision loss. One of the main causes of blindness and vision impairment in diabetics in industrialized countries is diabetes retinopathy (DR), a crippling chronic condition. Studies show that with early detection and treatment, the majority of cases may be averted. During eye screening, doctors use retinal imaging to find lesions related to this disease. The number of photos that need to be carefully reviewed is increasing because there are more and more diabetics. It also takes a while to train new employees in this type of image-based diagnosis because experience is gained via constant practice. In this paper, computer vision-based classification technique developed for identifying diabetic retinopathy from the retina fundus. The feature extraction method is based on a pretrained convolutional neural network which has rich set of feature layers. The classification is performed with a support vector machine algorithm that is based on machine learning. The proposed experimentation is evaluated and analyzed in MATLAB software with 96.66% accuracy and found to be efficient as compared to state of art methods.