app.py

import gradio as gr
import numpy as np
import cv2
import tensorflow as tf
from tensorflow.keras.preprocessing.image import ImageDataGenerator

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


# Preprocessing
test_datagen = ImageDataGenerator(rescale=1./255)

# Class labels
class_names = ['0.Anthracnose', '1.Fruit Fly', '2.Healthy Guava']

# Prediction function
def classify_image(image):
    """
    Process and classify the input image.
    Args:
        image: Input image in PIL format.

    Returns:
        Predicted class label.
    """
    # Convert to numpy array
    opencv_image = np.array(image)
    # Resize and preprocess the image
    img = cv2.resize(opencv_image, (150, 150))
    img = np.expand_dims(img, axis=0).astype('float32')  # Expand dimensions
    img = test_datagen.standardize(img)  # Normalize the image
    # Predict using the model
    predictions = model.predict(img)
    predicted_class = class_names[np.argmax(predictions)]
    return predicted_class

# Gradio Interface
interface = gr.Interface(
    fn=classify_image,
    inputs=gr.Image(type="pil", label="Upload an image"),
    outputs=gr.Textbox(label="Predicted Disease"),
    title="Guava Fruit Disease Classification",
    description=(
        "This app classifies diseases in Guava fruits using deep learning. "
        "Upload an image of a papaya to get started."
    ),
    
    allow_flagging="never",
)

# Launch the app
interface.launch()