app.py

import tensorflow as tf
import cv2
import numpy as np
import sqlite3
import gradio as gr

# ============================
# Load Model
# ============================

model = tf.keras.models.load_model("efficient_model.keras")

# ============================
# Load Database
# ============================

conn = sqlite3.connect("food_recipes.db", check_same_thread=False)
cursor = conn.cursor()

# ============================
# Classes
# ============================

class_names = [
    "chocolate_cake",
    "cup_cakes",
    "donuts",
    "dumplings",
    "french_fries",
    "fried_rice",
    "garlic_bread",
    "pizza",
    "samosa",
    "waffles",
]

# ============================
# Image Preprocessing
# ============================

def sq_img(img):
    h, w = img.shape[:2]

    if h >= w:
        diff = h - w
        left = diff // 2
        right = diff - left
        top = bottom = 0
    else:
        diff = w - h
        top = diff // 2
        bottom = diff - top
        left = right = 0

    return cv2.copyMakeBorder(
        img,
        top,
        bottom,
        left,
        right,
        cv2.BORDER_CONSTANT,
        value=0,
    )

# ============================
# Prediction Function
# ============================

def predict(image):

    if image is None:
        return "Please upload an image.", ""

    img = sq_img(image)
    img = cv2.resize(img, (224, 224))
    img = np.expand_dims(img, axis=0)

    probs = model.predict(img, verbose=0)

    pred = np.argmax(probs, axis=1)[0]
    confidence = float(np.max(probs)) * 100

    if confidence < 80:
        return (
            f"Prediction Confidence: {confidence:.2f}%\n\nPlease upload another image.",
            "",
        )

    food_name = class_names[pred]

    cursor.execute(
        "SELECT food_recipe FROM recipe WHERE name=?",
        (food_name,),
    )

    result = cursor.fetchone()

    recipe = result[0] if result else "Recipe not found."

    prediction = f"### 🍽️ {food_name.replace('_',' ').title()}\nConfidence : **{confidence:.2f}%**"

    return prediction, recipe


# ============================
# Gradio Interface
# ============================

demo = gr.Interface(
    fn=predict,
    inputs=gr.Image(type="numpy", label="Upload Food Image"),
    outputs=[
        gr.Markdown(label="Prediction"),
        gr.Markdown(label="Recipe"),
    ],
    title="🍞 Food Recipe Search",
    description="Upload a food image to identify it and get its recipe.",
)

if __name__ == "__main__":
    demo.launch(
        server_name="0.0.0.0",
        server_port=7860
    )