import gradio as gr
from tensorflow.keras.models import load_model
import numpy as np
from PIL import Image

model = load_model('xray_image_classifier_model.keras')

def predict(image):
    img = image.resize((150, 150))
    img_array = np.array(img) / 255.0
    img_array = np.expand_dims(img_array, axis=0)

    # Make a prediction
    prediction = model.predict(img_array)
    predicted_class = 'Pneumonia' if prediction > 0.5 else 'Normal'

    return predicted_class


css = """
    .gradio-container {
        background-color: #f5f5f5;
        font-family: Arial, sans-serif;
    }
    .gr-button {
        background-color: #007bff;
        color: white;
        border-radius: 5px;
        font-size: 16px;
    }
    .gr-button:hover {
        background-color: #0056b3;
    }
    .gr-textbox, .gr-image {
        border: 2px dashed #007bff;
        padding: 20px;
        border-radius: 10px;
        background-color: #ffffff;
    }
    .gr-box-text {
        color: #007bff;
        font-size: 22px;
        font-weight: bold;
        text-align: center;
    }
    h1 {
        font-size: 36px;
        color: #007bff;
        text-align: center;
    }
    p {
        font-size: 20px;
        color: #333;
        text-align: center;
    }
"""

# Gradio interface set up
with gr.Blocks(css=css) as interface:
    gr.Markdown("<h1>Chest X-ray Pneumonia Classifier</h1>")
    gr.Markdown("<p>Upload an X-ray image to classify it as 'Pneumonia' or 'Normal'.</p>")
    
    with gr.Row():
        image_input = gr.Image(label="Drop Image Here", type="pil", elem_classes=["gr-image", "gr-box-text"])
        output = gr.Textbox(label="Prediction", elem_classes=["gr-textbox", "gr-box-text"])

    submit_btn = gr.Button("Classify X-ray", elem_classes=["gr-button"])
    submit_btn.click(fn=predict, inputs=image_input, outputs=output)

interface.launch()