Feature extraction and classification of retinal images for automated detection of Diabetic Retinopathy

The disorders related to retina of the eye like Diabetic Retinopathy (DR), Age-related Macular Degeneration (AMD), and Glaucoma etc., can cause visual impairments. These disorders can be diagnosed by the ophthalmologists with the help of the Digital image processing. The retinal fundus images of the patients are procured by capturing the fundus of the eye with a digital fundus camera. The Automated method of disease detection can be used against the manual method of observing several retinal fundus images to save time. In this paper a method for DR detection by utilizing Fuzzy C-Means (FCM) clustering and morphological image processing is proposed. The image pre-processing includes image resizing, CLAHE, contrast adjustment, gray and green channel extraction from the color fundus image. The classification by Support Vector Machine (SVM) classifier using selected features achieves an Accuracy of 96.67%, Sensitivity of 100%, and Specificity of 95.83%.