A hybrid approach for diagnosing diabetic retinopathy from fundus image exploiting deep features

One of the major causes blindness in human beings is diabetic retinopathy (DR). To prevent blindness, early detection DR therefore necessary. In this paper, a hybrid model proposed for diagnosing from fundus images. A combination morphological image processing and Inception v3 deep learning techniques are exploited to detect as well classify healthy, mild non-proliferative (NPDR), moderate NPDR, severe proliferative (PDR). The algorithm was carried out several steps such segmentation blood vessels, localization removal optic disc, macula, abnormal features (microaneurysms, hemorrhages, neovascularization), classification. Microaneurysms hemorrhages that appear retina signs DR. work, we have detected microaneurysms by applying dynamic contrast limited adaptive histogram equalization threshold value on overlapping patched An overall accuracy 96.83% obtained into five different stages. better performance demonstrates effectiveness novelty work compared recent reported work.