Detection of exudates in human fundus image with a comparative study on methods for the optic disk detection

Retinopathy is a serious problem faced by people with diabetes. Once vision is lost because of this disease, usually it cannot be restored. Ophthalmologists clinically diagnose this by detecting the presence of exudates in the fundus images. Exudates are the deposition of the fluids leaked out from the damaged vessels in the retina. The effectiveness of the treatment depends upon the early detection of these exudates. Here, we propose a method for the automatic detection in the human fundus image by the application of the digital image processing. The methods adopted here were the optic disk detection using circular Hough transform and bit plane slicing, extraction of the exudates using morphological operations. Two methods adopted for the detection of optic disk were compared in this study. The proposed method is a novel approach. It has a sensitivity of 0.9362 and an accuracy of 88%.