Detection of Eye Ailments Using Segmentation of Blood Vessels from Eye Fundus Image

Eyes are vital part of the body which can be affected by many diseases that lead to vision loss. Glaucoma is one such eye disease that may cause vision loss. There are multiple reasons for vision loss which may be due to the appearance of unwanted blood vessels that can be caused by high level of glucose in the blood composition. This abnormal growth or change in behavior of the blood vessels represents underlying indicators of problems associated with eye diseases such as diabetic retinopathy. Hence, early detection of eye ailments can be expedited with the help of various image processing technologies. The first step after image acquisition is the processing of images to extract features that exactly match the disease under observation. This paper attempts to evaluate the blood vessels using different segmentation algorithms and introduce an improved version of the vessel algorithm. The evaluation of segmentation approaches shows that Otsu clustering algorithm is performing best as compared to other state-of-the-art techniques using eye fundus images.