Diabetic Retinopathy Detection and Classification Using Deep-wavelet and multiple classifiers

The eyes, being the most intricate and indispensable sensory organs, serve as a window for gathering external images which are transformed to signals and delivered to the brain through the optic nerves. There are many diseases can affect the human eye. Diabetic Retinopathy (DR) is one of these eye diseases especially affecting people with diabetes. It caused a thickening of the outer retinal layer and is known to have different levels of severity. In recent times, automatic classification of certain ophthalmologic and cardiovascular diseases using analysis of retinal images has become commonplace within telemedicine settings. In this paper we proposed a novel methodology in two stages for the classification and detection of Diabetic Retinopathy (DR) along with various severities of the disease using deep wavelet as a tool for feature extraction (FE) techniques with modalities such as KNN, SVD and for detection and classification tasks. The dataset used is the Diabetic Retinopathy Grayscale Images dataset from Kaggle .In our case, we conclude that our approach is performing reasonably and provides a good results according to the comparative estimate.