Detection of Glaucoma Disease using Image Processing, Soft Computing and Deep Learning Approaches

Glaucoma disease becomes a more common eye disease that occurs due to pressure on eye cells. Many image processing based methods have been applied earlier for the detection of glaucoma disease but their accuracy of classification was not up to the mark. The pressure on eye cells increases with the use of mobile phones, video games in the daily life of human beings. In this paper, the three different methods ares shared for the detection of glaucoma disease using image processing techniques, machine learning techniques, and using a convolutional neural network model of deep learning on the Bin Rushed database. The image processing techniques are used for the extraction of features like CDR and RDR, then classification performed using a neural network, support vector machine, decision tree, and K nearest machine learning model. The highest accuracy of 98% got for K nearest neighbor method and the VVG-16 deep learning model accuracy was 99.6%.