Diabetic Retinopathy Identification and Severity Classification

Manual examination of retina images for the diagnosis of diabetic retinopathy is a time consuming and error prone process, requiring identification of inconspicuous anomalies like micro-aneurysms and exudates. In this work, we explore machine learning techniques for automatic identification and severity classification of diabetic retinopathy from retina images. The presented approach involves image pre-processing, feature extraction using the bag of visual words model, and a multi-class classifier to classify the image into different DR stages. We have considered SURF, LBP and HoG features for constructing the bag of visual words. For the multi-class classification, we have implemented multinomial logistic regression, SVM and random forests.