import numpy as np
from flask import Flask, request, render_template, jsonify
from tensorflow.keras.models import load_model
from tensorflow.keras.preprocessing import image
from PIL import Image, ImageChops, ImageEnhance
import io
import base64

from util import base64_to_pil

app = Flask(__name__)

# Load the pre-trained model
model = load_model('model/model_vgg16_pretrained_80_100-90.h5')

def perform_ela(img, scale=10):
    ela_image = img.copy().convert('RGB')
    ela_image.save('temp_ela.jpg', 'JPEG', quality=90)
    ela_image = Image.open('temp_ela.jpg')

    diff = ImageChops.difference(img, ela_image)
    extrema = diff.getextrema()
    max_diff = max([ex[1] for ex in extrema])
    
    scale_factor = 255.0 / max_diff
    diff = ImageEnhance.Brightness(diff).enhance(scale_factor)

    return diff

def model_predict(img, model):
    img = img.resize((128, 128))
    x = image.img_to_array(img)
    x = np.expand_dims(x, axis=0)  # Menambah dimensi batch
    x = x.astype('float32') / 255.0
    preds = model.predict(x)
    return preds

@app.route('/', methods=['GET'])
def index():
    return render_template('index.html')

@app.route('/predict', methods=['POST'])
def predict():
    try:
        if request.method == 'POST':
            # Get the image from POST request
            img = base64_to_pil(request.json)

            # Perform ELA
            ela_img = perform_ela(img)

            # Make prediction using ELA image
            preds = model_predict(ela_img, model)

            # Konversi prediksi dari nilai probabilitas menjadi kelas (0 atau 1)
            pred_class = np.argmax(preds, axis=1)[0]

            # Lakukan pengembalian hasil prediksi
            if pred_class == 0:
                hasil_label = 'Manipulasi'
            else:
                hasil_label = 'Real'

            hasil_prob = "{:.2f}".format(100 * np.max(preds))

            return jsonify(result=hasil_label, probability=hasil_prob)
        else:
            return jsonify({'error': 'Only POST requests are accepted'})
    except Exception as e:
        return jsonify({'error': str(e)})

if __name__ == '__main__':
    app.run(debug=True, use_reloader=False)