import numpy as np
import tensorflow as tf
import gradio as gr
from PIL import Image
from io import BytesIO
import matplotlib.pyplot as plt
import os 

# Load the trained TensorFlow model using the Keras API
model = tf.keras.models.load_model("potatoesV3.h5")

# Define the class names for classification
CLASS_NAMES = ["Early Blight", "Late Blight", "Healthy"]


# Define the function for making predictions

def classify_image(request):
    # Open the file using requests
    image = request.files["file"]

    image = np.array(
        Image.open(image).convert("RGB").resize((256, 256)) # image resizing
    )

    image = image/255 # normalize the image in 0 to 1 range

    img_array = tf.expand_dims(img, 0)
    predictions = model.predict(img_array)

    print("Predictions:",predictions)

    predicted_class = class_names[np.argmax(predictions[0])]
    confidence = round(100 * (np.max(predictions[0])), 2)

    return {"class": predicted_class, "confidence": confidence}
    # Resize the image to the desired size
    # new_size = (256, 256)
    # img_resized = image.resize(new_size)
    
    # # Convert the image to a numpy array
    # img_array = np.array(img_resized)
    
    # # Rescale the image pixel values to [0, 1]
    # img_scaled = img_array / 255.0

    # #img_array = np.array(img_scaled)
    # img_array = np.expand_dims(img_scaled, axis=0)
    # pred = model.predict(img_array)
    
    # # Display the rescaled image
    # #plt.imshow(img_array)
    
    # # Format the predicted class and confidence score
    # predicted_class = CLASS_NAMES[np.argmax(pred[0])]
    # confidence = float(np.max(pred[0]))

    # # Return the predicted class and confidence score
    # return {"class": predicted_class, "confidence": confidence,"predict":pred[0]}
    # img = (BytesIO(image))
    # img = img.resize((256, 256))
    # img_array = np.array(img)
    # img_array = np.expand_dims(img_array, axis=0)
    # pred = model.predict(img_array)
    # return pred[0][0]

examples=[os.path.join(os.path.dirname(__file__),'7227b3db-c212-4370-8b42-443eea1577aa___RS_Early.B 7306.JPG','7456db33-766c-4a68-b924-ddf69d579981___RS_Early.B 6723.JPG','7486e823-64f7-4e43-ab51-26261b077fc2___RS_Early.B 6785.JPG','8829e413-5a7a-4680-b873-e71dfa9dbfe4___RS_LB 3974.JPG','9001b18c-b659-4c56-9dfb-0d0bf64a7b4a___RS_LB 4987.JPG','9009c86e-1205-4694-b0bb-ef7cf78dd104___RS_LB 3995.JPG','Potato_healthy-76-_0_2420.jpg','Potato_healthy-76-_0_6833.jpg','Potato_healthy-76-_0_7539.jpg')]

# Define the Gradio interface for image input
input_interface = gr.inputs.File(label="Upload an image file")

# Define the Gradio interface for displaying the predicted class and confidence score
output_interface = gr.Textbox()

# Launch the Gradio interface with the classify_image function as the prediction function
gr.Interface(fn=classify_image, inputs=input_interface, outputs=output_interface,Examples=examples).launch()