A Study and Comparison of Automated Techniques for Exudate Detection Using Digital Fundus Images of Human Eye: A Review for Early Identification of Diabetic Retinopathy

Exudates are a visible sign of diabetic retinopathy which is the major cause of blindness in patients with diabetes. If the exudates extend into the macular area, vision loss can occur. Automated early detection of the presence of exudates can assist ophthalmologists to prevent the spread of the disease more efficiently. Hence, detection of exudates is an important diagnostic task. Exudates are found using their high grey level variation. The detection of the optic disc is indispensable in the exudates detection process since they both are similar in terms of color, contrast, etc. Here a study of various techniques like morphological approach, region growing approach, fuzzy c-means clustering technique, k-means clustering techniques is made to suggest a technique for automatic early identification of diabetic retinopathy. These four techniques are implemented and their performances are evaluated based on various metrics like sensitivity, specificity, etc. These algorithms are tested on a small image data base and their accuracies are analyzed with respect to expert ophthalmologist’s hand-drawn ground-truths.