Data_Preparation.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "provenance": [],
      "collapsed_sections": [],
      "authorship_tag": "ABX9TyPzMc0GrBGagUgDwi4QR6dP",
      "include_colab_link": true
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    },
    "language_info": {
      "name": "python"
    },
    "gpuClass": "standard"
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "view-in-github",
        "colab_type": "text"
      },
      "source": [
        "<a href=\"https://colab.research.google.com/github/lorenzrossi/MachineLearning-CatsDogs-NN/blob/master/Data_Preparation.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "L8cP3Kq5ohHr",
        "outputId": "a918d268-f936-454a-d46f-f78a20d3243a"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Mounted at /content/drive\n",
            "24978\n",
            "0\n",
            "0\n",
            "0\n",
            "1\n",
            "0\n",
            "0\n",
            "0\n",
            "1\n",
            "0\n"
          ]
        }
      ],
      "source": [
        "import os\n",
        "from os import listdir\n",
        "import pandas as pd\n",
        "# Basic packages\n",
        "import numpy as np\n",
        "import matplotlib.pyplot as plt\n",
        "import seaborn as sns\n",
        "import random\n",
        "import pickle\n",
        "\n",
        "# Image preprocessing\n",
        "import pathlib\n",
        "import PIL\n",
        "import PIL.Image\n",
        "import cv2\n",
        "\n",
        "from google.colab import drive\n",
        "drive.mount('/content/drive')\n",
        "\n",
        "dir = '/content/drive/MyDrive/CatsDogs'\n",
        "\n",
        "# Import of Data\n",
        "\n",
        "# I define some parameters in order resize image in an uniform way anf I set 3 as number of channels of the image, since I want RGB images\n",
        "img_width = 100\n",
        "img_height = 100\n",
        "img_size = (img_width, img_height)\n",
        "channels = 1\n",
        "#%%\n",
        "# Pets is the list in which alle the images will be stored\n",
        "pets = []\n",
        "\n",
        "# I define the two categories of pets ( related to the sub-directories)\n",
        "categories = ['Cats','Dogs']\n",
        "\n",
        "# I create a funciotn in order to load the data, transform the into array, assign them to a class and store them in the pets list\n",
        "def create_data():\n",
        "\n",
        "# the function iterates throught the two sub-directories\n",
        "    for category in categories:\n",
        "        path = os.path.join(dir, category)\n",
        "\n",
        "# assign 0 to cat and 1 to dog, according to the index of the categories\n",
        "        pet_class = categories.index(category)\n",
        "\n",
        "# the function then iterates through each image in both the folders. I used the opencv (cv2 when imported) package to read and load them. IMREAD_COLOR keeps the color in the image, whereas IMREAD_GRAYSCALE would turn them into grayscale images.\n",
        "# The function resize the images as defined by the previous parameters and finally appends the array of the image, together with the class value (creating a sub-list of two values for each image), to the pets list.\n",
        "# The function skips those images which seem broken.\n",
        "        for img in os.listdir(path):\n",
        "            try:\n",
        "                img_array = cv2.imread(os.path.join(path,img), cv2.IMREAD_GRAYSCALE)\n",
        "                #img_array = cv2.bitwise_not(img_array)\n",
        "                new_array = cv2.resize(img_array, img_size)\n",
        "                pets.append([new_array, pet_class])\n",
        "            except:\n",
        "                pass\n",
        "\n",
        "create_data()\n",
        "\n",
        "print(len(pets))\n",
        "\n",
        "# since the previous function iterates linearly through the two sub-directories, iìI shuffle the data in order to mix shuffle the order of the images\n",
        "random.shuffle(pets)\n",
        "\n",
        "for sample in pets[:9]:\n",
        "    print(sample[1])\n",
        "\n",
        "# split the data arrays and the labels for the classification task\n",
        "\n",
        "X = []\n",
        "y = []\n",
        "\n",
        "for image, label in pets:\n",
        "    X.append(image)\n",
        "    y.append(label)\n",
        "\n",
        "X = np.array(X).reshape(-1, img_width, img_height, channels)\n",
        "\n",
        "pickle_out = open('/content/drive/MyDrive/CatsDogs/Pickles/X.pickle','wb')\n",
        "pickle.dump(X, pickle_out)\n",
        "pickle_out.close()\n",
        "\n",
        "pickle_out = open('/content/drive/MyDrive/CatsDogs/Pickles/y.pickle','wb')\n",
        "pickle.dump(y, pickle_out)\n",
        "pickle_out.close()\n",
        "\n",
        "pickle_in = open('/content/drive/MyDrive/CatsDogs/Pickles/X.pickle','rb')\n",
        "X = pickle.load(pickle_in)\n",
        "\n",
        "pickle_in = open('/content/drive/MyDrive/CatsDogs/Pickles/y.pickle','rb')\n",
        "y = pickle.load(pickle_in)"
      ]
    },
    {
      "cell_type": "code",
      "source": [],
      "metadata": {
        "id": "36XOqjzxHNgB"
      },
      "execution_count": null,
      "outputs": []
    }
  ]
}