imageClassification.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "id": "60ecfa3b-ade0-4a44-829e-7a875265ef89",
   "metadata": {},
   "source": [
    "# IMAGE CLASSIFICATION\n",
    "## Top Secret Rosies & HP Challenge 2023\n",
    "### By Anna Gracia Colmenarejo\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "f5e9c751-3c25-471d-bdc4-bebc35300e9f",
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "from sklearn.neighbors import KNeighborsClassifier\n",
    "from sklearn.metrics import confusion_matrix, ConfusionMatrixDisplay, classification_report\n",
    "import cv2\n",
    "import os\n",
    "import matplotlib.pyplot as plt\n",
    "from sklearn.linear_model import SGDClassifier\n",
    "from sklearn.preprocessing import StandardScaler\n",
    "from sklearn.pipeline import make_pipeline"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "855c75db",
   "metadata": {},
   "source": [
    "##### This function reads all the images in the folder. These images are separated by training images and test images. The training images will be used to train the classification algorithm and the test images will be used to test the algorithm and classify them. Finally, for the training and test data structures, we will save each photo with its respective class."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "ecb04e56-6d22-42c3-9fcc-86bdcb539d41",
   "metadata": {},
   "outputs": [],
   "source": [
    "def load_images(path, clases):\n",
    "    Xtr = []\n",
    "    Ytr = []\n",
    "    Xte = []\n",
    "    Yte = []\n",
    "    for i in clases:\n",
    "        tr = path+i+'/train/'\n",
    "        ts = path+i+'/test/'\n",
    "        for img in os.listdir(tr):\n",
    "            img = cv2.imread(os.path.join(tr, img))\n",
    "            img_resized = cv2.resize(img,(200,400))\n",
    "            Xtr.append(img_resized)\n",
    "            Ytr.append(clases.index(i))\n",
    "        for img in os.listdir(ts):\n",
    "            img = cv2.imread(os.path.join(ts, img))\n",
    "            img_resized = cv2.resize(img,(200,400))\n",
    "            Xte.append(img_resized)\n",
    "            Yte.append(clases.index(i))\n",
    "    # Convert list to numpy array format\n",
    "    Xtr = np.array(Xtr)\n",
    "    Ytr = np.array(Ytr)\n",
    "    Xte = np.array(Xte)\n",
    "    Yte = np.array(Yte)\n",
    "\n",
    "    return Xtr, Ytr, Xte, Yte"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "6afd3eac",
   "metadata": {},
   "outputs": [],
   "source": [
    "def knn_algorithm(k, x_train, y_train, x_test, y_test):\n",
    "    knn = KNeighborsClassifier(n_neighbors=k)\n",
    "    knn.fit(x_train, y_train)\n",
    "    y_pred = knn.predict(x_test) \n",
    "    print(\"Prediction complete\")\n",
    "    print('The accuracy of classifier on test data: {:.2f}'.format(knn.score(x_test, y_test)*100))\n",
    "    return y_pred"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "774311e5",
   "metadata": {},
   "outputs": [],
   "source": [
    "def SGD_algorithm(x_train, y_train, x_test, y_test):\n",
    "    clf = make_pipeline(StandardScaler(), SGDClassifier(max_iter=1000, tol=1e-3))\n",
    "    clf.fit(x_train, y_train)\n",
    "    y_pred = clf.predict(x_test)\n",
    "    print(\"Prediction complete\")\n",
    "    print('The accuracy of classifier on test data: {:.2f}'.format(clf.score(x_test, y_test)*100))\n",
    "    return y_pred"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "ee6aa191",
   "metadata": {},
   "source": [
    "##### We load the training and test images and their clases using our function load_images(images_dir,clases)."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "8a00049c",
   "metadata": {},
   "outputs": [],
   "source": [
    "images_dir = 'Images/'\n",
    "clases = ['Documents','Plans','Tickets']\n",
    "x_train, y_train, x_test, y_test = load_images(images_dir, clases)"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "9678d958",
   "metadata": {},
   "source": [
    "##### Reshape the image data into rows."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "091f5ac3-8f27-42c9-babf-f1552dc3e199",
   "metadata": {},
   "outputs": [],
   "source": [
    "x_train = np.reshape(x_train, (x_train.shape[0], -1))\n",
    "x_test = np.reshape(x_test, (x_test.shape[0], -1))"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "6fac6c60",
   "metadata": {},
   "source": [
    "##### First we will be using KNN algorithm"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "f4a48db6",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Prediction complete\n",
      "The accuracy of classifier on test data: 90.00\n",
      "Test set classification report\n",
      "\n",
      "              precision    recall  f1-score   support\n",
      "\n",
      "           0       1.00      0.99      1.00       181\n",
      "           1       0.90      0.83      0.86       203\n",
      "           2       0.80      0.88      0.84       156\n",
      "\n",
      "    accuracy                           0.90       540\n",
      "   macro avg       0.90      0.90      0.90       540\n",
      "weighted avg       0.90      0.90      0.90       540\n",
      "\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<sklearn.metrics._plot.confusion_matrix.ConfusionMatrixDisplay at 0x1e9db109390>"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 800x800 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "y_pred_knn = knn_algorithm(5, x_train, y_train, x_test, y_test)\n",
    "print(\"Test set classification report\\n\")\n",
    "print(classification_report(y_test, y_pred_knn))\n",
    "\n",
    "cm = confusion_matrix(y_test,y_pred_knn)\n",
    "display = ConfusionMatrixDisplay(cm, display_labels=clases)\n",
    "\n",
    "fig, ax = plt.subplots(figsize=(8,8))\n",
    "display.plot(ax=ax, values_format='')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "f10ef397-a8c7-4a8c-8e1f-db9c42942fcd",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Prediction complete\n",
      "The accuracy of classifier on test data: 94.44\n",
      "Test set classification report\n",
      "\n",
      "              precision    recall  f1-score   support\n",
      "\n",
      "           0       0.98      1.00      0.99       181\n",
      "           1       0.93      0.94      0.93       203\n",
      "           2       0.93      0.89      0.91       156\n",
      "\n",
      "    accuracy                           0.94       540\n",
      "   macro avg       0.94      0.94      0.94       540\n",
      "weighted avg       0.94      0.94      0.94       540\n",
      "\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<sklearn.metrics._plot.confusion_matrix.ConfusionMatrixDisplay at 0x1e9db62a110>"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 800x800 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "y_pred_sgd = SGD_algorithm(x_train, y_train, x_test, y_test)\n",
    "print(\"Test set classification report\\n\")\n",
    "print(classification_report(y_test, y_pred_sgd))\n",
    "\n",
    "cm = confusion_matrix(y_test,y_pred_sgd)\n",
    "display = ConfusionMatrixDisplay(cm, display_labels=clases)\n",
    "\n",
    "fig, ax = plt.subplots(figsize=(8,8))\n",
    "display.plot(ax=ax, values_format='')"
   ]
  }
 ],
 "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": "3.11.2"
  },
  "vscode": {
   "interpreter": {
    "hash": "c261aea317cc0286b3b3261fbba9abdec21eaa57589985bb7a274bf54d6cc0a7"
   }
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}