Multi Categorical of Common Eye Disease Detect Using Convolutional Neural Network

Abstract

The eye is one of the most vital organs in the human body. Humans, despite their diminutive stature, are unable to see life without it. A thin covering known as the conjunctiva protects the human eye from dust particles. It works as a lubricant in the eye, preventing friction during the eye's opening and closing. There are various different types of eye ailments. Because the human eye is the most important of the four sense organs, it is necessary to detect external eye diseases early. Pattern recognition can be applied to a wide range of scenarios. One of these apps is a medical application. In this paper we use Convolutional Neural Network different architecture to detect normal eyes, conjunctivitis eyes and cataract eyes where apply ResNet50, VGG 16 and Inception v3. Among them ResNet50 performs 99 percent accuracy to detect eye disease with 485s time taken to detect. Respectively, Inception v3 performs 97 percent accuracy and VGG16 performs 95 percent accuracy.