Preprocessing.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import os\n",
    "import cv2\n",
    "import keras\n",
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "from sklearn.utils import shuffle\n",
    "from keras.preprocessing.image import load_img, img_to_array\n",
    "from keras.utils import to_categorical\n",
    "import glob"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Augmenting the Healthy___Potato\n",
    "Previously the no. of images for the Healthy potato are : 152 . After **Augmentation** it becomes 1008"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "# from keras.preprocessing.image import ImageDataGenerator, array_to_img, img_to_array, load_img\n",
    "\n",
    "# datagen = ImageDataGenerator(rotation_range =15, \n",
    "#                          width_shift_range = 0.2, \n",
    "#                          height_shift_range = 0.2,  \n",
    "#                          rescale=1./255, \n",
    "#                          shear_range=0.2, \n",
    "#                          zoom_range=0.2, \n",
    "#                          horizontal_flip = True, \n",
    "#                          fill_mode = 'nearest') \n",
    "\n",
    "\n",
    "# img_dir = \"Potato/Potato___healthy/\" \n",
    "# data_path = os.path.join(img_dir,'*jpg')\n",
    "# files = glob.glob(data_path)\n",
    "# data = []\n",
    "# for f1 in files:\n",
    "#     img = cv2.imread(f1)\n",
    "#     data.append(img)\n",
    "    \n",
    "# print(len(data))\n",
    "\n",
    "# x = np.array(data)\n",
    "# x = x.reshape(x.shape)\n",
    "\n",
    "# i = 0\n",
    "\n",
    "# for batch in datagen.flow (x,\n",
    "#                            batch_size=32,\n",
    "#                            save_to_dir ='Potato/Heathy___potato/',\n",
    "#                            save_prefix=\"a\",\n",
    "#                            save_format='jpg'):\n",
    "#     i+=1\n",
    "#     if i==33:\n",
    "#       break"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Reading the data from the directory itself:\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(256, 256, 3)\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "DIR = \"D:/research/Projects/Potato plant disease/Potato\"\n",
    "CATEGORIES = ['Potato___Early_blight', 'Heathy___potato', 'Potato___Late_blight']\n",
    "\n",
    "\n",
    "for category in CATEGORIES:\n",
    "    path = os.path.join(DIR, category)\n",
    "    for image in os.listdir(path):\n",
    "        img_arr = cv2.imread(os.path.join(path, image))\n",
    "        plt.imshow(img_arr)\n",
    "        print(img_arr.shape)\n",
    "        break\n",
    "    break\n",
    "    \n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Making the training data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(256, 256, 3)\n",
      "(0,)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n",
      "Shape:  (224, 224, 3)\n"
     ]
    }
   ],
   "source": [
    "training_data = []\n",
    "\n",
    "def create_training_data():\n",
    "    for category in CATEGORIES:\n",
    "        path = os.path.join(DIR, category)\n",
    "        classes = CATEGORIES.index(category)\n",
    "        \n",
    "        for image in os.listdir(path):\n",
    "            img = load_img(os.path.join(path, image), target_size=(224, 224))        \n",
    "            img_arr = img_to_array(img)\n",
    "            img_arr = img_arr/255\n",
    "            print(\"Shape: \", img_arr.shape)\n",
    "            training_data.append([img_arr, classes])\n",
    "    \n",
    "print(img_arr.shape)\n",
    "    \n",
    "img_data = np.array(training_data)\n",
    "img_data = img_data.astype('float32')\n",
    "print (img_data.shape)\n",
    "\n",
    "            \n",
    "create_training_data()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "3008\n"
     ]
    }
   ],
   "source": [
    "print(len(training_data))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Shuffling the data:\n",
    "\n",
    "- `0` : Early blight potato\n",
    "- `1` : Healthy potato\n",
    "- `2` : Late blight potato\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0\n",
      "1\n",
      "1\n",
      "0\n",
      "1\n",
      "2\n",
      "0\n",
      "2\n",
      "1\n",
      "1\n"
     ]
    }
   ],
   "source": [
    "import random\n",
    "\n",
    "random.shuffle(training_data)\n",
    "\n",
    "for sample in  training_data[:10]:\n",
    "    print(sample[1])\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "features = []\n",
    "labels = []\n",
    "\n",
    "\n",
    "for x, y in training_data:\n",
    "    features.append(x)\n",
    "    labels.append(y)\n",
    "    \n",
    "x = np.array(x).reshape(-1, 224, 224, 3)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "3008\n"
     ]
    }
   ],
   "source": [
    "print(len(labels))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pickle\n",
    "\n",
    "# dumping the features and labels\n",
    "pickle_out = open(\"features.pkl\",\"wb\")\n",
    "pickle.dump(features, pickle_out)\n",
    "pickle_out.close()\n",
    "\n",
    "pickle_out = open(\"labels.pkl\",\"wb\")\n",
    "pickle.dump(labels, pickle_out)\n",
    "pickle_out.close()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.8.10"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}