Glaucoma is eye diseases that result into total vision loss and blindness. It can be prevented by early detection. Glaucoma has no early symptoms or feeling of pain in eye at the early stage of Glaucoma, thus it is required to see the doctor regularly as early detection and treatment will prevent further loss of vision. Proposed methodology based on convolution neural network that is useful for differentiating normal image and glaucoma image. There are various other features as well as which can be used for the classification of glaucoma. Thus, appropriate feature selection becomes important in order to improve the efficiency without degrading the performance of the classifier. Out of various features which can be used for the classification of glaucoma not all are useful, also redundant features can reduce the performance. In this paper we develop a model adding additional layer in the U-Net architecture in order to gain better accuracy.
Volume 11 | 04-Special Issue
Pages: 2543-2554