# -*- coding: utf-8 -*- # Import required library from __future__ import division, print_function # coding=utf-8 import base64 import io import json import sys import os import glob import re import numpy as np import cv2 import matplotlib.pyplot as plt import matplotlib.patches as patches # Keras from PIL import Image from tensorflow.keras.applications.imagenet_utils import preprocess_input, decode_predictions from tensorflow.keras.models import load_model from tensorflow.keras.preprocessing import image # Flask utils from flask import Flask, redirect, url_for, request, render_template from werkzeug.utils import secure_filename from gevent.pywsgi import WSGIServer # Define a flask app # app = Flask(__name__) app = Flask(__name__) # Model saved with Keras model.save() dependencies = { 'IoU': '' } # Add model path # mobilenet model accuracy 80% to 85% in between # MODEL_PATH = 'my_model_23_03.h5' # MODEL_PATH = 'my_model_22_03_1.h5' MODEL_PATH = 'model.h5' # Load your trained model # model = load_model(MODEL_PATH) # model = load_model(MODEL_PATH, custom_objects=dependencies) # model._make_predict_function() # Necessary def model_predict_car(img_path, model): # print('in function predict car ',selected_model.sel_model) # # if(selected_model.sel_model=='Mobilenet'): # print('if Mobilenet') # elif(selected_model.sel_model=='Resnet 50'): # print('else Resnet 50') # elif(selected_model.sel_model=='VGG16'): # print('else VGG16') filename = img_path image_size = 128 unscaled = cv2.imread(filename) # print(unscaled) image_height, image_width, _ = unscaled.shape image = cv2.resize(unscaled, (image_size, image_size)) # Rescaled image to run the network feat_scaled = preprocess_input(np.array(image, dtype=np.float32)) region = model.predict(x=np.array([feat_scaled]))[0] # Predict the BBox print('region ', region) x0 = int(region[0] * image_width / image_size) # Scale the BBox y0 = int(region[1] * image_height / image_size) x1 = int((region[0] + region[2]) * image_width / image_size) y1 = int((region[1] + region[3]) * image_height / image_size) # Create figure and axes fig, ax = plt.subplots(1) # Display the image ax.imshow(unscaled) # Create a Rectangle patch rect = patches.Rectangle((x0, y0), x1 - x0, y1 - y0, linewidth=2, edgecolor='r', facecolor='none') # Add the patch to the Axes ax.add_patch(rect) plt.show() def model_predict_car_new(img_path, model): filename = img_path image_size = 128 unscaled = cv2.imread(filename) # print(unscaled) # print('filename ', filename) # plt.imshow(cv2.imread(filename)) # plt.imshow(unscaled) image_height, image_width, _ = unscaled.shape image = cv2.resize(unscaled, (image_size, image_size)) # Rescaled image to run the network feat_scaled = preprocess_input(np.array(image, dtype=np.float32)) region_1 = model.predict(x=np.array([feat_scaled])) # Predict the BBox print(region_1) region=region_1[0] print('region ', region) x0 = int(region[0] * image_width / image_size) # Scale the BBox y0 = int(region[1] * image_height / image_size) x1 = int((region[0] + region[2]) * image_width / image_size) y1 = int((region[1] + region[3]) * image_height / image_size) # Blue color in BGR color = (255, 0, 0) # Line thickness of 2 px thickness = 2 # cv2.rectangle(unscaled, (x0, y0), (x1 + x0, y1 + y0),color,thickness) cv2.rectangle(unscaled, (x0, y0), (x1, y1), color, thickness) # # Create figure and axes # fig, ax = plt.subplots(1) # # # Display the image # ax.imshow(unscaled) # # # # Create a Rectangle patch # # rect = patches.Rectangle((x0, y0), x1 - x0, y1 - y0, linewidth=2, edgecolor='r', facecolor='none') # # # # # Add the patch to the Axes # # ax.add_patch(rect) # # # plt.show() return unscaled @app.route('/', methods=['GET']) def index(): # Main page return render_template('home.html') @app.route('/home', methods=['GET']) def home(): # Main page return render_template('home.html') @app.route('/selected_model', methods=['POST']) def selected_model(): selected_model.sel_model = request.form['sel_model']; print('_sel_model ', selected_model.sel_model) return json.dumps({'status': 'OK', 'selected_model': selected_model.sel_model}); from PIL import Image as im @app.route('/predict', methods=['GET', 'POST']) def upload(): if request.method == 'POST': # Get the file from post request f = request.files['file'] # Save the file to ./uploads basepath = os.path.dirname(__file__) file_path = os.path.join( basepath, 'uploads', secure_filename(f.filename)) f.save(file_path) # Make prediction # preds = model_predict_car(file_path, model) preds = model_predict_car_new(file_path, model) print(preds.shape) # print() print('Predict image') # print(preds) # Process your result for human ## pred_class = preds.argmax(axis=-1) # Simple argmax # pred_class = decode_predictions(preds, top=1) # ImageNet Decode # result = str(pred_class[0][0][1]) # Convert to string # return result data = im.fromarray(preds) print('file name ' + f.filename) # Delete old file files = glob.glob('static/save_img/*') for f1 in files: os.remove(f1) # files = glob.glob('images/*') # for f1 in files: # os.remove(f1) k = f.filename.rfind(".") # printing the filename file_name=f.filename[:k]; print('f.filename ' +f.filename[:k]) # print('k ' +k) # saving the final output # as a PNG file a = 'static/save_img/' b = file_name c = '.png' file_p = a + b + c data.save(file_p) # np_img = Image.fromarray(preds) # print('from array',np_img) # img_encoded = image_to_byte_array(np_img) # print('Encode image',img_encoded) # base64_bytes = base64.b64encode(img_encoded).decode("utf-8") # # return jsonify({'status': True, 'image': image}) print('return image') return render_template('show_image.html', user_image=file_p) # return None if __name__ == '__main__': app.run(debug=True)