app.py

from flask import Flask, render_template, request, redirect, url_for, session
from flask_session import Session
import tensorflow as tf
import cv2

import numpy as np
import os
import matplotlib.pyplot as plt

from tensorflow.keras.models import load_model
model_keras = load_model("generator_model.h5")

app = Flask(__name__)
app.config['SESSION_TYPE'] = 'filesystem' 
Session(app)

SIZE =128
def image_upload(path):
    images = []
    images.append(path)
    return np.array(images)

def convert_image_inputs(images):
    labels = []
    for image_path in images:
        img = cv2.imread(image_path)
        if img is not None:
            img = cv2.cvtColor(img, cv2.COLOR_BGR2GRAY)
            # res_img = cv2.resize(img, (SIZE, SIZE), interpolation=cv2.INTER_CUBIC)
            res_img = cv2.resize(img, (SIZE, SIZE), interpolation=cv2.INTER_LINEAR)
            labels.append(res_img)
    return labels, img.shape[1], img.shape[0]


def make_color(path_input):
    image = image_upload(path_input)
    input_img, w, h = convert_image_inputs(image)
    input_img = np.array(input_img)
    input_img = input_img/255.0
    generated_image = model_keras.predict(input_img)
    return generated_image, w, h

def result(output):
    num_samples = len(output)
    fig = plt.figure()
    ax = fig.add_subplot(1, 1, 1)
    ax.imshow(output)
    ax.set_title('Output')
    plt.show()


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

@app.route('/upload', methods=['GET','POST'])
def upload_image():
    if request.method == 'POST':
        if 'image' in request.files:
            image = request.files['image']
            if image.filename != '':
                image_path = os.path.join('static', 'uploads', image.filename)
                image.save(image_path)

                generated_image, img_width, img_height = make_color(image_path)
                print(generated_image[0], img_width, img_height)
                generated_image = generated_image[0]
                generated_image = generated_image*255.0
                
 
                generated_image_filename = 'generated_' + image.filename
                generated_image_path = os.path.join('static', 'generated', generated_image_filename)
                generated_image_rgb = cv2.cvtColor(generated_image, cv2.COLOR_BGR2RGB)
                # SIZE = 1080
                res_img1 = cv2.resize(generated_image_rgb, (img_height, img_width), interpolation=cv2.INTER_CUBIC)
                cv2.imwrite(generated_image_path, res_img1)
                session['before_image_url'] = url_for('static', filename='uploads/' + image.filename)
                session['after_image_url'] = url_for('static', filename='generated/' + generated_image_filename)
                return redirect(url_for('show_image', filename=generated_image_filename))

    return render_template('result.html')

@app.route('/show/<filename>')
def show_image(filename):
    uploaded_image_url = url_for('static', filename='generated/' + filename)
    before_image_url = session.get('before_image_url', '')
    after_image_url = session.get('after_image_url', '')
    return render_template('result.html', uploaded_image_url=uploaded_image_url,  before_image_url=before_image_url, after_image_url=after_image_url)

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