import gradio as gr
import tensorflow as tf
from tensorflow.keras.models import load_model
import numpy as np
from tensorflow.keras.preprocessing import image



def predict_input_image(img):
    # Normalize the image by cropping (center crop)
    h, w = img.shape[:2]
    crop_start_x = (w - 224) // 2
    crop_start_y = (h - 224) // 2
    img = img[crop_start_y:crop_start_y+224, crop_start_x:crop_start_x+224]
    img = tf.image.resize(img, [224,224])
    img = np.expand_dims(img, axis = 0)
    
    # Make predictions
    model = tf.keras.models.load_model('Tumor_Model.h5')
    prediction = model.predict(img)
    result = 'No Tumor Detected' if prediction[0][0] > 0.5 else 'Tumor detected'
    

    return prediction

    

# Define Gradio interface
iface = gr.Interface(
    fn=predict_input_image,
    inputs= 'image',
    outputs="text",
)

# Launch the interface
iface.launch()