Direct Cup-to-Disc Ratio Estimation for Glaucoma Screening via Semi-supervised Learning

Abstract

A series of connected eye conditions known as glaucoma harm the optic nerve, which transmits information from the eye to the brain, and can eventually result in blindness. For proper treatment, it is crucial that glaucoma is identified as soon as possible. A Convolutional Neural Network (CNN) approach for the early identification of Glaucoma is suggested in this research. To start, augmented ocular pictures are used to produce data for deep learning. The ocular pictures are then pre-processed using the Gaussian Blur technique to reduce noise and prepare the image for additional processing. When new input images are provided to the system, it classifies them as either normal eyes or glaucoma eyes based on the features collected during training. The system is trained using the pre-processed images.