Load_and_Resize.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Image loading and resizing"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**Importing Libraries**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 53,
   "metadata": {},
   "outputs": [],
   "source": [
    "import os\n",
    "import numpy as np\n",
    "import pandas as pd\n",
    "import matplotlib.pyplot as plt\n",
    "\n",
    "from PIL import Image\n",
    "from tqdm.notebook import tqdm\n",
    "from IPython.display import display, HTML"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Lego dataset"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 67,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Define the path to the lego dataset DATA directory\n",
    "data_path = os.getcwd() + '/DATA/'"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Some things to notice about the dataset:\n",
    "- It is formed by 200x200 pixel images.\n",
    "- They are in .PNG format: supports 32-bit RGBA colors and has the ability to use transparency and or opacity, as opposed to JPG where by default it will retain a white/blank background.\n",
    "- We have 6379 different images in the dataset: 15 classes with 400 images and one with 379."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Here is an example:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 57,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(-0.5, 199.5, 199.5, -0.5)"
      ]
     },
     "execution_count": 57,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x432 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Take one image at random\n",
    "im = data_path + '2357 Brick corner 1x2x2/201706171206-0048.png'\n",
    "\n",
    "image = Image.open(im).convert('L')\n",
    "image = np.asarray(image).astype('uint8')\n",
    "\n",
    "fig, ax = plt.subplots(figsize = (6, 6))\n",
    "ax.imshow(image, cmap = 'gray', vmin = 0, vmax = 255)\n",
    "ax.axis('off')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Now we will read the whole dataset and store it in a simple DataFrame:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 66,
   "metadata": {},
   "outputs": [],
   "source": [
    "def read_all_data(path):\n",
    "    ''' Function that takes a path to a directory and reads all the images and stores them as a \n",
    "        pandas dataframe.\n",
    "    ----------\n",
    "    PARAMETERS\n",
    "    - path: string indicating the full path of the parent directory\n",
    "    ----------\n",
    "    RETURNS\n",
    "    - df200, df128, df64: pd.DataFrame with all the flattened grayscale images read\n",
    "    \n",
    "    '''\n",
    "    # Loop through all the class labels\n",
    "    for it, label in enumerate(os.listdir(data_path)):\n",
    "        # Beware of hidden files in the directory\n",
    "        if label[0] != '.':\n",
    "            # Full path to the different sub-directories\n",
    "            label_dir = data_path + label + '/'\n",
    "            \n",
    "            for i, im in enumerate(os.listdir(label_dir)):\n",
    "                \n",
    "                try:\n",
    "                    # Open the image and convert the PNG to grayscale\n",
    "                    image = Image.open(label_dir + im).convert('L')\n",
    "                    \n",
    "                    # Normal 200x200 resolution (no scaling)\n",
    "                    im200 = list(np.asarray(image).flatten().astype('uint8'))\n",
    "                    im200.append(label)\n",
    "                    \n",
    "                    # Downsized 128x128 resolution\n",
    "                    im128 = image.resize((128, 128), Image.ANTIALIAS)\n",
    "                    im128 = list(np.asarray(im128).flatten().astype('uint8'))\n",
    "                    im128.append(label)\n",
    "                    \n",
    "                    # Downsized 64x64 resolution\n",
    "                    im64 = image.resize((64, 64), Image.ANTIALIAS)\n",
    "                    im64 = list(np.asarray(im64).flatten().astype('uint8'))\n",
    "                    im64.append(label)\n",
    "                    \n",
    "                    # Create variables if first iteration\n",
    "                    if i == 0:\n",
    "                        m200 = im200\n",
    "                        m128 = im128\n",
    "                        m64 = im64\n",
    "                    # Else append it to the matrix\n",
    "                    else:\n",
    "                        m200 = np.vstack((m200, im200))\n",
    "                        m128 = np.vstack((m128, im128))\n",
    "                        m64 = np.vstack((m64, im64))\n",
    "                \n",
    "                except:\n",
    "                    print(f'Failed reading image: {im}')\n",
    "            \n",
    "            # Create pd.DataFrame variable if first iteration\n",
    "            if it == 0: \n",
    "                df200 = pd.DataFrame(m200)\n",
    "                df128 = pd.DataFrame(m128)\n",
    "                df64 = pd.DataFrame(m64)\n",
    "            # Else append it to the existing dataframe\n",
    "            else: \n",
    "                df200 = df200.append(pd.DataFrame(m200))\n",
    "                df128 = df128.append(pd.DataFrame(m128))\n",
    "                df64 = df64.append(pd.DataFrame(m64))\n",
    "                \n",
    "    return (df200, df128, df64)      "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 68,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "CPU times: user 1h 23min 40s, sys: 42min 20s, total: 2h 6min 1s\n",
      "Wall time: 32min 53s\n"
     ]
    }
   ],
   "source": [
    "%%time\n",
    "\n",
    "df200, df128, df64 = read_all_data(data_path)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Take a look at the dataset (the 64x64 one for example):"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 69,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>0</th>\n",
       "      <th>1</th>\n",
       "      <th>2</th>\n",
       "      <th>3</th>\n",
       "      <th>4</th>\n",
       "      <th>5</th>\n",
       "      <th>6</th>\n",
       "      <th>7</th>\n",
       "      <th>8</th>\n",
       "      <th>9</th>\n",
       "      <th>...</th>\n",
       "      <th>4087</th>\n",
       "      <th>4088</th>\n",
       "      <th>4089</th>\n",
       "      <th>4090</th>\n",
       "      <th>4091</th>\n",
       "      <th>4092</th>\n",
       "      <th>4093</th>\n",
       "      <th>4094</th>\n",
       "      <th>4095</th>\n",
       "      <th>4096</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>3005 Brick 1x1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>3005 Brick 1x1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>3005 Brick 1x1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>3005 Brick 1x1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>3005 Brick 1x1</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 4097 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "  0    1    2    3    4    5    6    7    8    9     ... 4087 4088 4089 4090  \\\n",
       "0    0    0    0    0    0    0    0    0    0    0  ...    0    0    0    0   \n",
       "1    0    0    0    0    0    0    0    0    0    0  ...    0    0    0    0   \n",
       "2    0    0    0    0    0    0    0    0    0    0  ...    0    0    0    0   \n",
       "3    0    0    0    0    0    0    0    0    0    0  ...    0    0    0    0   \n",
       "4    0    0    0    0    0    0    0    0    0    0  ...    0    0    0    0   \n",
       "\n",
       "  4091 4092 4093 4094 4095            4096  \n",
       "0    0    0    0    0    0  3005 Brick 1x1  \n",
       "1    0    0    0    0    0  3005 Brick 1x1  \n",
       "2    0    0    0    0    0  3005 Brick 1x1  \n",
       "3    0    0    0    0    0  3005 Brick 1x1  \n",
       "4    0    0    0    0    0  3005 Brick 1x1  \n",
       "\n",
       "[5 rows x 4097 columns]"
      ]
     },
     "execution_count": 69,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df64.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Hence we have 4096 columns representing the 64x64 flattened pixels of every image, and the last column representing the label of the image."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Finally, store the dataframes into csv files in your working directory:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {},
   "outputs": [],
   "source": [
    "export200 = df200.to_csv(r'df200.csv', index = False, header = False)\n",
    "export128 = df128.to_csv(r'df128.csv', index = False, header = False)\n",
    "export64 = df64.to_csv(r'df64.csv', index = False, header = False)"
   ]
  }
 ],
 "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": "3.6.8"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}