Input Image Dataset (e.g., Fundus Images)

Preprocessing

Image Cleaning (Noise Reduction)

Normalization (Intensity Scaling)

Resizing and Cropping

Data Augmentation (Rotation, Flipping, Zoom)

Deep Learning Model (CNN)

Model Architecture Selection

Transfer Learning (Pre-trained CNNs like VGG, ResNet)

Custom CNN Layers (Convolutional, Pooling, Activation)

Model Training (Optimization, Loss Function)

Hyperparameter Tuning

Feature Extraction

Classification

Output: Disease Diagnosis (e.g., Diabetic Retinopathy, Glaucoma)

Database (Optional) - Processed Data, Models, Results