Feature extraction for early detection of macular hole and glaucoma in fundus images

Retinal fundus image acquired with digital fundus cameras is adaptable tools for the diagnosis of common retinal diseases. Digital Image Processing is extensively used in present biomedical applications for feature detection and classification of diseases. In this paper, the focus is on fundus photographs to detect the features of two common retinal diseases, namely, macular hole and glaucoma using the preprocessing algorithms and feature extraction algorithms of Digital Image Processing. The classification of these two diseases into their different stages is not in the scope of this research work. An early detection of such diseases enables the ophthalmologist to educate the patient about the progress of the disease and attend such patients immediately for the further diagnosis. The studies have also shown that these diseases being treated in their early stages yield a better result compared to being treated in their later stages.