mnist_preprocessing.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "from tensorflow.keras.datasets import mnist\n",
    "import matplotlib.pyplot as plt\n",
    "import numpy as np\n",
    "from skimage.transform import resize"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "(train_image, train_gt), (test_image, test_gt) = mnist.load_data()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "def cut(image):\n",
    "    \"\"\"crops the image so that no row or column is blank\"\"\"\n",
    "    result = []\n",
    "    for row in image:\n",
    "        if set(list(row)) != set([0]): result.append(row)\n",
    "    return result"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "def crop(image):\n",
    "    \"\"\"crops all four edges using the cut method\"\"\"\n",
    "    return np.array(cut(np.array(cut(image)).transpose())).transpose()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\MYChe\\Anaconda3\\lib\\site-packages\\skimage\\transform\\_warps.py:105: UserWarning: The default mode, 'constant', will be changed to 'reflect' in skimage 0.15.\n",
      "  warn(\"The default mode, 'constant', will be changed to 'reflect' in \"\n",
      "C:\\Users\\MYChe\\Anaconda3\\lib\\site-packages\\skimage\\transform\\_warps.py:110: UserWarning: Anti-aliasing will be enabled by default in skimage 0.15 to avoid aliasing artifacts when down-sampling images.\n",
      "  warn(\"Anti-aliasing will be enabled by default in skimage 0.15 to \"\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<matplotlib.image.AxesImage at 0x2374f529b70>"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "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": [
    "#testing method on image from the MNIST dataset\n",
    "\n",
    "image = train_image[930]\n",
    "new_image = crop(image)\n",
    "result = resize(new_image, (15,15))\n",
    "plt.imshow(result)"
   ]
  }
 ],
 "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": "3.6.6"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}