import streamlit as st
import streamlit.web.cli as stcli
import tensorflow as tf
import numpy as np
from PIL import Image

IMAGE_SIZE = 256

# Load the saved model
model = tf.keras.models.load_model('my_model.h5')

# Define class labels (adjust this according to your specific classes)
class_labels = ['Mild', 'Moderate', 'No_DR', 'Proliferate_DR', 'Severe']


def predict(image):
    # Preprocess the image to the required size and scale
    image = tf.image.resize(image, (IMAGE_SIZE, IMAGE_SIZE))
    image = np.expand_dims(image, axis=0)  # Add batch dimension

    # Make prediction
    predictions = model.predict(image)
    confidence = np.max(predictions)
    predicted_class = class_labels[np.argmax(predictions)]

    return predicted_class, float(confidence)


# Create the Streamlit interface
st.title("Early Diabetic Retinopathy Detection")
st.write("Upload an image and get the predicted class along with confidence score.")

uploaded_file = st.file_uploader("Choose an image...", type=["jpg", "jpeg", "png"])

if uploaded_file is not None:
    image = Image.open(uploaded_file)
    st.image(image, caption='Uploaded Image.', use_column_width=True)
    st.write("")
    st.write("Classifying...")

    predicted_class, confidence = predict(image)

    st.write(f"Predicted Class: {predicted_class}")
    st.write(f"Confidence: {confidence:.2f}")