This project involves building a Glaucoma Detection AI-ML model using a Convolutional Neural Network (CNN) to classify retinal images as either "Glaucoma Affected" or "Normal." The model is trained using ImageDataGenerator for data augmentation, with binary cross-entropy loss, Adam optimizer, and is saved in .keras format. The user can upload an image, and the model predicts glaucoma presence with a percentage, suggesting medical consultation if affected.
IMPORTANT NOTE : In the uploaded Train and Validation folder, here I only uploaded some sample photos. But in original you have to upload as much photos as you can collect (only in .jpg/.jpeg) format. If I tell about mine, In actual I used more than 7000 photos for train and more than 1000 photos for validation for the better accuracy.
The Website User Interface :-
If the uploaded retinal image Glaucoma Affected, Then :-
If the uploaded retinal image Normal, Then :-
