Analysis and Diagnosis of Glaucoma

Glaucoma is a disease which damages the optic nerve, the part of the eye which carries theimages in the form of electrical impulses to the brain, and leads to loss of vision. It is causeddue to the immense increase in the intraocular pressure in which rate of secretion of the ciliary body is unbalanced to the rate of drainage in the drainage canal. Depending on how the pressure rises, glaucoma is classified as Open angle and Angle-closure glaucoma. At later stages of life, glaucoma carmot be treated due to severe damage to the optic nerve as disease prolongs, therefore, early stages of glaucoma detection are necessary. Manual detection of glaucoma is difficult and very sensitive, which depends on the specialist. The present work provides an image processing technique for prediction of glaucoma. Preprocessing is performed on the retinal fundus image by eliminating the blood vessels in the retinal fundus image to equalize the irregularities present in the image. Feature extraction is done by evaluating cup to disc ratio using ellipse fitting algorithm and color component analysis technique is applied on the input retinal fundus image to determine to evaluate the area of cup to disc which determines abnormal image. Classification is done using Bayes classifier to isolate glaucomatous images from healthy images.