import streamlit as st
from tensorflow.keras.models import load_model
from tensorflow.keras.preprocessing.image import img_to_array  # Import this function
from PIL import Image
import numpy as np
import os
from huggingface_hub import hf_hub_download, login

# Title of the Streamlit app
st.title("Yellow Rust Severity Prediction")

# Authentication using Hugging Face token
authkey = os.getenv('YellowRust')
login(token=authkey)

# Download the model file from Hugging Face
model_path = hf_hub_download(repo_id="shaheer-data/Yellow-Rust-Prediction", filename="final_meta_model.keras")

# Load the pre-trained model
loaded_model = load_model(model_path)

# Function to preprocess the uploaded image
def preprocess_image(image):
    # Resize the image to match the model input size (e.g., 224x224 for many pre-trained models)
    image = image.resize((224, 224))  # Adjust size based on your model input
    image = img_to_array(image)       # Convert image to numpy array
    image = image / 255.0            # Normalize pixel values to [0, 1]
    image = np.expand_dims(image, axis=0)  # Add batch dimension
    return image

# Streamlit file uploader
uploaded_file = st.sidebar.file_uploader("Upload a wheat leaf image", type=["jpg", "jpeg", "png"])

if uploaded_file is not None:
    st.sidebar.subheader("Uploaded Image")
    image = Image.open(uploaded_file)
    st.sidebar.image(image, caption="Uploaded Image", use_container_width=True)

    # Preprocess the image
    processed_image = preprocess_image(image)

    st.subheader("Prediction: With 97% Accuracy")
    # Perform prediction
    with st.spinner("Predicting..."):
        prediction = loaded_model.predict(processed_image)
        predicted_class = np.argmax(prediction, axis=1)[0]  # Get the class index
        class_labels = ['0', 'MR', 'MRMS', 'MS', 'R', 'S']  # Update based on your classes
        
        st.header("Predicted Severity Class")
        
        if predicted_class == 0:
            st.markdown('<p style="color: green; font-size: 20px;">Healthy</p>', unsafe_allow_html=True)
            st.write("The leaf appears healthy. There is no immediate action required. Continue monitoring as needed.")
        elif predicted_class == 1:
            st.markdown('<p style="color: orange; font-size: 20px;">Mild Rust (MR)</p>', unsafe_allow_html=True)
            st.write("Mild rust detected. Applying fungicides will help control further spread.")
        elif predicted_class == 2:
            st.markdown('<p style="color: #FFA500; font-size: 20px;">Moderate Rust (MRMS)</p>', unsafe_allow_html=True)
            st.write("Moderate rust detected. Monitor regularly and treat with fungicides.")
        elif predicted_class == 3:
            st.markdown('<p style="color: #FF4500; font-size: 20px;">Severe Rust (MS)</p>', unsafe_allow_html=True)
            st.write("Severe rust detected. Prompt fungicide application and continued monitoring are recommended.")
        elif predicted_class == 4:
            st.markdown('<p style="color: red; font-size: 20px;">Very Severe Rust (R)</p>', unsafe_allow_html=True)
            st.write("Very severe rust detected. Intensive control measures and frequent monitoring are required.")
        elif predicted_class == 5:
            st.markdown('<p style="color: darkred; font-size: 20px;">Extremely Severe Rust (S)</p>', unsafe_allow_html=True)
            st.write("Extremely severe rust detected. Apply aggressive control strategies and seek expert advice.")
        
        confidence = np.max(prediction) * 100
        st.write(f"**Confidence Level:** {confidence:.2f}%")

# Footer
st.info("MPHIL Final Year Project By Mr. Asim Khattak")