Classification of IRIS Recognition Based on Deep Learning Techniques

Abstract

In this work, a man-made awareness instruct passes on an extremely profitable sort of significant making sense of how to investigate eye sickness from helpful pictures. Convolution Neural systems are significant learning computations capable of getting ready pictures and predicting the diseases. The output will be determined as either Conjunctiva and Corneal ulcer according to the decision made based on the estimated features from the image. CNNs give the best execution in model and picture affirmation issues, even outmaneuvering humans in certain aspects. An ensemble of network architecture improved prediction accuracy. An independent dataset was used to evaluate the performance of our algorithm in a population-based study. Class conjunctivitis is labeled as 0, and Corneal Ulcer is classified as 1. After extracting the features, model is classifying the image as 0 or 1. To further improve the model, this paper is used VGG-16, ResNet 152 and Alexnet to predict the disease and improve accuracy.