Diabetic Retinopathy Detection Using Voting Classification Method

Abstract

The image processing approach is deployed to process the information available in the pixels of images. The medical images processing method is extensively exploited to detect distinct types of diseases. The retinal disease may occur in the diabetic patients that leads to create specific spots around the eyes of an infected individual. This disorder is called diabetic retinopathy. The image processing-based technique (IPT) are effective to diagnose such a disease in 2 stages. First of all, the features are extracted and the disease is classified later on. The present work suggests a voting classification technique which is the combination of SVM, KNN and random forest for the diabetic retinopathy detection. The suggested method is tested concerning accuracy, precision and recall.