Comparative Study of Methods Used for Diagnosis of Diabetic Retinopathy Using Image Processing

India is the world’s second most populated country this fact also contributes to the verity that why India is known as the Diabetes capital of the world. According to IDF (or International Diabetic Federation), on an average 463 million people around the globe are suffering from diabetes out of which 77 million alone resides in India. By scientists around the world it is also estimated that India will reach a benchmark of 134 million diabetes cases, just the double amount it is now by next 25 years. In India diabetes, is popularly known as “sugar” or “sugar level”. And several studies have concluded a point that basically “Type 2” diabetes is more common in India, which causes Indians a greater risk towards “Diabetic Retinopathy (or popularly known as DR)”. DR is a medical condition which arises in a person if he/she is a patient of diabetes for at least 10 years. Although in its earlier stage, DR can be cured under the supervision of the doctor, but if a person neglects the warnings in its initial stages, then this disease becomes uncontrollable in its later stages even leading to the vision impairment by permanently destroying the retinal vessels of the patient’s eye. This research paper directs towards the different machine learning models that can be used in the field of medical science for early detection of DR which can help to detect this disease at the initial stage using Fundus images of patients and can further help for the betterment of our society, nation and the overall world.