Sclera Feature Extraction for Detection of Diabetic Disease

Abstract: Sclera is a white and opaque outer protective covering of the eye. It is stable over the life time and unique to each person. A very few researchers have studied about the color changesoccur in sclera, but low accuracy, initial results. This is the challenging to identify the particular disease. Here segmentation is performed to separate the sclera from the image; k-means clustering is used. Enhancement is necessary to develop the accurate vessel structure for that Gabor filter is used. To detect the disease it is important to extract the feature for the identification of the particular disease. Here LBP method is used to extract the feature of the sclera. SVM is used to classify the affected and non-affected sclera regions.

Keywords: Diabetes, Segmentation, Sclera, LBP, SVM classifier.