Regresar

Portable Nuclear and Cortical Eye Cataract Detection using Image Processing

Abstract:

Early detection and treatment of cataract can prevent worse effects in eyesight, such as blurred vision and blindness. However, portability is an issue with the machines that are used for detecting eye cataracts. The focus of this paper will be the detection and classification of cataract using image processing algorithms which are Local Ridge Enhancement, Sobel Algorithm, Directional Coding, and Back Propagation Neural Network. The mobile application that shall serve as this study's output will display whether the user has an eye cataract or not. If an eye cataract is detected, the app shall only display the type of cataract and the severity grade of the eye cataract. Based on the data gathered and the doctor's assessment, this paper resulted in a 100% accuracy on detecting and classifying Normal Eyes, 80.00% on Nuclear Mild Cataract, 85.00% on Nuclear Severe Cataract, 80.00% on Cortical Mild Cataract, and 93.33% on Cortical Severe Cataract.