shape-classification-hough-lbp-svm-rf-99.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "id": "49124cd6",
   "metadata": {
    "papermill": {
     "duration": 0.003218,
     "end_time": "2023-11-09T22:27:46.746478",
     "exception": false,
     "start_time": "2023-11-09T22:27:46.743260",
     "status": "completed"
    },
    "tags": []
   },
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "e43dc7ed",
   "metadata": {
    "_cell_guid": "b1076dfc-b9ad-4769-8c92-a6c4dae69d19",
    "_uuid": "8f2839f25d086af736a60e9eeb907d3b93b6e0e5",
    "execution": {
     "iopub.execute_input": "2023-11-09T22:27:46.754327Z",
     "iopub.status.busy": "2023-11-09T22:27:46.753610Z",
     "iopub.status.idle": "2023-11-09T22:32:44.220327Z",
     "shell.execute_reply": "2023-11-09T22:32:44.219014Z"
    },
    "papermill": {
     "duration": 297.47429,
     "end_time": "2023-11-09T22:32:44.223680",
     "exception": false,
     "start_time": "2023-11-09T22:27:46.749390",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "/kaggle/input/four-shapes/shapes/circle\n",
      "/kaggle/input/four-shapes/shapes/square\n",
      "/kaggle/input/four-shapes/shapes/star\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAgMAAAELCAYAAABEYIWnAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjcuMywgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy/OQEPoAAAACXBIWXMAAA9hAAAPYQGoP6dpAABEJElEQVR4nO3deXhTVf4/8PfN0rRJmqQrtKUtq9CWogICsgoifBVQHFdABEXFfUXccByXH446OvqoKK7gggygIiqioKBsCrLIvreldKNb2rRJm+38/mCSITRpQ2mbtvf9ep4+ys1p8sltcu/nnnvO50hCCAEiIiKSLUWoAyAiIqLQYjJAREQkc0wGiIiIZI7JABERkcwxGSAiIpI5JgNEREQyx2SAiIhI5pgMEBERyRyTASIiIpljMtCE/vGPf0CSpEb97oIFCyBJErKzs5s2qNNkZ2dDkiQsWLCg2V6DiAgApk+fjs6dO4c6DL9xSJKEf/zjHy0eS6heNxhMBgDs3bsXN910E5KSkqDRaJCYmIgpU6Zg7969oQ4tJNatWwdJkrBs2bJQh0LtmCcB/vPPP/0+fskll6B3797n9BrffvstJkyYgA4dOiAsLAzR0dEYPnw4Xn31VVRWVjb4+9OnT4ckSX5/Vq1adU6xBbJo0SK8/vrrzfLcjeVwOBAbG4uhQ4cGbCOEQHJyMvr27duCkbU+K1eubLUn/PqoQh1AqH311VeYNGkSoqOjMWPGDHTp0gXZ2dn48MMPsWzZMixevBhXX311UM81Z84cPP74442KY+rUqbjxxhuh0Wga9ftE9D9utxszZszAggULkJmZibvvvhvJycmwWCzYvHkz5syZg5UrV+Lnn39u8Lk0Gg0++OCDOtvPP//85ggdixYtwp49e/Dggw82y/M3hlqtxnXXXYf58+cjJycHqampddr89ttvOHHiBB566CEAwPvvvw+3293SoQbFZrNBpWqe09/KlSvx9ttv+00ImvN1z1XrjKqFHD16FFOnTkXXrl3x22+/IS4uzvvYAw88gGHDhmHq1KnYtWsXunbtGvB5qqurodPpoFKpGv2HViqVUCqVjfpdIvL18ssvY8GCBXjooYfw6quv+ty+e+CBB1BQUIBPPvkkqOdSqVS46aabmivUFmO1WqHVahv9+1OmTMG7776LL774wu9Fz6JFi6BQKHDjjTcCOJVAtFbh4eGyet1gyPo2wSuvvAKr1Yr33nvPJxEAgNjYWMyfPx/V1dV4+eWXvds94wL27duHyZMnIyoqytt15m/MgM1mw/3334/Y2FhERkbiyiuvRF5eXp17R/7GDHTu3Bnjx4/Hhg0bMGDAAISHh6Nr1651DmJlZWWYNWsWMjMzodfrYTAYcPnll+Ovv/5qoj31v/d26NAh3HTTTTAajYiLi8PTTz8NIQRyc3Nx1VVXwWAwoGPHjnj11Vd9ft9ut+Pvf/87+vXrB6PRCJ1Oh2HDhmHt2rV1Xqu0tBRTp06FwWCAyWTCtGnT8Ndff/kd73DgwAFce+21iI6ORnh4OPr3748VK1Y02fum1sXpdOL5559Ht27doNFo0LlzZzz55JOora31trFarXjppZeQkZGBV155xe84noSEBDz22GNNEpPb7cbrr7+OjIwMhIeHo0OHDpg5cybKy8t92n3zzTcYN24cEhMTodFo0K1bNzz//PNwuVzeNpdccgm+//575OTkeG9HeO53BxpX5Lmtt27dOp/n6d27N7Zt24bhw4dDq9XiySefBADU1tbimWeeQffu3aHRaJCcnIzZs2f77EN/hgwZgs6dO2PRokV1HnM4HFi2bBlGjhyJxMREAP7v1S9evBj9+vVDZGQkDAYDMjMz8cYbb3gfDzTuyt97D2Z/BnL68dczlirQj8f69etx3XXXISUlxbvfHnroIdhsNm+b6dOn4+233/a+xpnP4W/MwI4dO3D55ZfDYDBAr9fj0ksvxe+//+73/W/cuBEPP/ww4uLioNPpcPXVV6O4uLjB9xsMWfcMfPvtt+jcuTOGDRvm9/Hhw4ejc+fO+P777+s8dt1116FHjx6YO3cu6lsFevr06ViyZAmmTp2KQYMG4ddff8W4ceOCjvHIkSO49tprMWPGDEybNg0fffQRpk+fjn79+iEjIwMAcOzYMSxfvhzXXXcdunTpgqKiIsyfPx8jRozAvn37vF/OpnDDDTcgLS0N//znP/H999/jhRdeQHR0NObPn49Ro0bhpZdewueff45Zs2bhoosuwvDhwwEAlZWV+OCDDzBp0iTcfvvtsFgs+PDDDzF27Fhs2bIFF1xwAYBTB9YJEyZgy5YtuOuuu9CrVy988803mDZtWp1Y9u7diyFDhiApKQmPP/44dDodlixZgokTJ+LLL78M+vYOhVZFRQVKSkrqbHc4HHW23XbbbVi4cCGuvfZaPPLII/jjjz/w4osvYv/+/fj6668BABs2bIDZbMasWbOarLftzPjUajWMRiMAYObMmViwYAFuueUW3H///cjKysJbb72FHTt2YOPGjd4r5AULFkCv1+Phhx+GXq/HL7/8gr///e+orKzEK6+8AgB46qmnUFFRgRMnTuDf//43AECv1zcq5tLSUlx++eW48cYbcdNNN6FDhw5wu9248sorsWHDBtxxxx1IS0vD7t278e9//xuHDh3C8uXLAz6fJEmYPHky5s6di71793qPPwCwatUqlJWVYcqUKQF/f/Xq1Zg0aRIuvfRSvPTSSwCA/fv3Y+PGjXjggQfO+v0Fsz+DERcXh08//dRnm8PhwEMPPYSwsDDvtqVLl8JqteKuu+5CTEwMtmzZgjfffBMnTpzA0qVLAZz6LOTn52P16tV1ntOfvXv3YtiwYTAYDJg9ezbUajXmz5+PSy65BL/++isGDhzo0/6+++5DVFQUnnnmGWRnZ+P111/Hvffei//85z9Bv9+AhEyZzWYBQFx11VX1trvyyisFAFFZWSmEEOKZZ54RAMSkSZPqtPU85rFt2zYBQDz44IM+7aZPny4AiGeeeca77eOPPxYARFZWlndbamqqACB+++0377aTJ08KjUYjHnnkEe+2mpoa4XK5fF4jKytLaDQa8dxzz/lsAyA+/vjjet/z2rVrBQCxdOnSOu/tjjvu8G5zOp2iU6dOQpIk8c9//tO7vby8XERERIhp06b5tK2trfV5nfLyctGhQwdx6623erd9+eWXAoB4/fXXvdtcLpcYNWpUndgvvfRSkZmZKWpqarzb3G63GDx4sOjRo0e975FCz/OZr+8nIyPD237nzp0CgLjtttt8nmfWrFkCgPjll1+EEEK88cYbAoBYvny5Tzun0ymKi4t9ftxud70xTps2zW9cI0aMEEIIsX79egFAfP755z6/t2rVqjrbrVZrneefOXOm0Gq1Pp/hcePGidTU1ID76/RjhBD/+76uXbvWu23EiBECgHj33Xd92n766adCoVCI9evX+2x/9913BQCxcePG+naH2Lt3rwAgnnjiCZ/tN954owgPDxcVFRXebdOmTfN5Hw888IAwGAzC6XQGfP4zj6Ee/t57sPvzzDiEEHWOv2e6++67hVKp9H6mAr3eiy++KCRJEjk5Od5t99xzj9/34O91J06cKMLCwsTRo0e92/Lz80VkZKQYPny4d5vn/Y8ePdrnM/vQQw8JpVIpzGZzwPcSLNneJrBYLACAyMjIett5Hj9z5PGdd97Z4Gt4RhvffffdPtvvu+++oONMT0/36bmIi4tDz549cezYMe82jUYDheLUn9LlcqG0tBR6vR49e/bE9u3bg36tYNx2223e/1cqlejfvz+EEJgxY4Z3u8lkqhOjUqn0ZtlutxtlZWVwOp3o37+/T4yrVq2CWq3G7bff7t2mUChwzz33+MRRVlaGX375Bddffz0sFgtKSkpQUlKC0tJSjB07FocPH0ZeXl6TvndqHm+//TZWr15d56dPnz4+7VauXAkAePjhh322P/LIIwDg7cHzfFfPvKLevXs34uLifH5KS0sbjC88PLxObJ7bYEuXLoXRaMRll13m/QyWlJSgX79+0Ov1PrfBIiIivP/v+cwOGzYMVqsVBw4cCGpfnQ2NRoNbbrnFZ9vSpUuRlpaGXr16+cQ7atQoAPB72+506enpuPDCC7F48WLvturqaqxYsQLjx4+HwWAI+LsmkwnV1dVYvXr1Obyr/2mu/fnJJ59g3rx5ePnllzFy5Ei/r1ddXY2SkhIMHjwYQgjs2LHjrF/H5XLhp59+wsSJE33GpCUkJGDy5MnYsGFDnfPOHXfc4XPbYdiwYXC5XMjJyTnr1z+TbG8TeE7ynqQgkEBJQ5cuXRp8jZycHCgUijptu3fvHnScKSkpdbZFRUX53I90u9144403MG/ePGRlZfncM4uJiQn6tRoTj9FoRHh4OGJjY+tsP/NAu3DhQrz66qs4cOCATxfw6fsnJycHCQkJdQY6nbnPjhw5AiEEnn76aTz99NN+Yz158iSSkpKCf3MUEgMGDED//v3rbI+KivLpnvd8n878LHTs2BEmk8l7QPR8V6uqqnzade/e3Xsi+uSTT4LqxgVOJbKjR4/2+9jhw4dRUVGB+Ph4v4+fPHnS+/979+7FnDlz8Msvv9Q5yFdUVAQVy9lISkry6eb2xLt///46Y6T8xRvIlClTMGvWLGzatAmDBw/G8uXLYbVa671FAJy6KFqyZAkuv/xyJCUlYcyYMbj++uvxf//3f8G/qdM0x/7cuXMn7rzzTkyaNKlO0nn8+HH8/e9/x4oVK+qMB2nM6xUXF8NqtaJnz551HktLS4Pb7UZubq7P7Zgzj79RUVEAUCeexpBtMmA0GpGQkIBdu3bV227Xrl1ISkqqk/GeniU2p0D3PMVp4xTmzp2Lp59+Grfeeiuef/55REdHQ6FQ4MEHH2zyqT3+4gkmxs8++wzTp0/HxIkT8eijjyI+Ph5KpRIvvvgijh49etZxeN7XrFmzMHbsWL9tzibporajocJevXr1AgDs2bMHV111lXe7Xq/3ntQ3bNjQJLG43W7Ex8fj888/9/u456RrNpsxYsQIGAwGPPfcc+jWrRvCw8Oxfft2PPbYY0F9TwO970AD5vwdo9xuNzIzM/Haa6/5/Z3k5OQG45g0aRJmz56NRYsWYfDgwVi0aBGioqJwxRVX1Pt78fHx2LlzJ3788Uf88MMP+OGHH/Dxxx/j5ptvxsKFCwEE/x6bYn+eqby8HNdccw3OO++8OlNJXS4XLrvsMpSVleGxxx5Dr169oNPpkJeXh+nTp7fYFMpgjrWNJdtkAADGjx+P999/Hxs2bPBbTGP9+vXIzs7GzJkzG/X8qampcLvdyMrKQo8ePbzbjxw50uiY/fGM4v3www99tpvN5jpX7KGybNkydO3aFV999ZXPF/6ZZ57xaZeamoq1a9fWmQZ15j7zdKup1eqAV23Uvni+T4cPH0ZaWpp3e1FREcxms3fu+7Bhw2A0GrF48WI88cQT3ltozaFbt25Ys2YNhgwZUu8Fwrp161BaWoqvvvrKO6gWALKysuq0DXRC9FwFms1mn+1n00XcrVs3/PXXX7j00ksbXS01MTERI0eOxNKlS/H0009j9erVmD59ep1eCH/CwsIwYcIETJgwAW63G3fffTfmz5+Pp59+Gt27d/d5jyaTyft7Z77Hs9mfwXC73ZgyZQrMZjPWrFlTp2dy9+7dOHToEBYuXIibb77Zu93fLY9g92tcXBy0Wi0OHjxY57EDBw5AoVAElZw1FdmOGQCARx99FBEREZg5c2adLu2ysjLceeed0Gq1ePTRRxv1/J4r1nnz5vlsf/PNNxsXcABKpbJOZrh06dJWdc/ck9GeHucff/yBzZs3+7QbO3YsHA4H3n//fe82t9vtna7jER8fj0suuQTz589HQUFBnddrquk21Hp4rjzPrM7nucr1zNLRarWYPXs29uzZg8cff9zvVVNTXEkBwPXXXw+Xy4Xnn3++zmNOp9N74vb3+bfb7XWODQCg0+n8djt369YNwKniPh4ulwvvvffeWcWbl5fn8/3ysNlsqK6uDup5pkyZgpMnT2LmzJlwOBwN3iIAUOcYq1AovONCPNMa/b3H6upqb8+Bx9nsz2A8++yz+PHHH/HFF1/4vQXs7/WEED7TIj10Oh2Aukmbv+ccM2YMvvnmG58pk0VFRVi0aBGGDh1a7xiMpibrnoEePXpg4cKFmDJlCjIzM+tUICwpKcEXX3zh/YCerX79+uGaa67B66+/jtLSUu/UwkOHDgEIPoNsyPjx4/Hcc8/hlltuweDBg7F79258/vnn9RZKamnjx4/HV199hauvvhrjxo1DVlYW3n33XaSnp/vc2504cSIGDBiARx55BEeOHEGvXr2wYsUKlJWVAfDdZ2+//TaGDh2KzMxM3H777ejatSuKioqwefNmnDhxoknrLFDonX/++Zg2bRree+89bzfxli1bsHDhQkycONFnsNfjjz+O/fv345VXXsFPP/2Ea665Bp06dUJ5eTm2b9+OpUuXIj4+/pyLwIwYMQIzZ87Eiy++iJ07d2LMmDFQq9U4fPgwli5dijfeeAPXXnstBg8ejKioKEybNg33338/JEnCp59+6jcp6devH/7zn//g4YcfxkUXXQS9Xo8JEyYgIyMDgwYNwhNPPIGysjJER0dj8eLFcDqdQcc7depULFmyBHfeeSfWrl2LIUOGwOVy4cCBA1iyZAl+/PFHv+M3znTNNdfg7rvvxjfffIPk5GSfq/NAbrvtNpSVlWHUqFHo1KkTcnJy8Oabb+KCCy7w9vSMGTMGKSkpmDFjBh599FEolUp89NFHiIuLw/Hjx73PdTb7syG7d+/G888/j+HDh+PkyZP47LPPfB6/6aab0KtXL3Tr1g2zZs1CXl4eDAYDvvzyS7/36vv16wcAuP/++zF27FgolUpvIaYzvfDCC1i9ejWGDh2Ku+++GyqVCvPnz0dtba1PfZsWcc7zEdqBXbt2iUmTJomEhAShVqtFx44dxaRJk8Tu3bvrtPVMfSkuLg742Omqq6vFPffcI6Kjo4VerxcTJ04UBw8eFAB8puMFmlo4bty4Oq8zYsQI79QmIU5NLXzkkUdEQkKCiIiIEEOGDBGbN2+u064pphae+b6nTZsmdDqd3xhPnxbmdrvF3LlzRWpqqtBoNOLCCy8U3333nd9pP8XFxWLy5MkiMjJSGI1GMX36dLFx40YBQCxevNin7dGjR8XNN98sOnbsKNRqtUhKShLjx48Xy5Ytq/c9Uuh5PvNbt271+/iZnyEhhHA4HOLZZ58VXbp0EWq1WiQnJ4snnnjCZyrZ6b7++mtxxRVXiLi4OKFSqYTJZBJDhw4Vr7zySlDTsQJ9vs/03nvviX79+omIiAgRGRkpMjMzxezZs0V+fr63zcaNG8WgQYNERESESExMFLNnzxY//vhjnWmBVVVVYvLkycJkMgkAPt+Po0ePitGjRwuNRiM6dOggnnzySbF69Wq/UwvP3HcedrtdvPTSSyIjI0NoNBoRFRUl+vXrJ5599lmfqYENue666wQAMXv2bL+Pn/ndXrZsmRgzZoyIj48XYWFhIiUlRcycOVMUFBT4/N62bdvEwIEDvW1ee+01v8fHYPdnQ1MLPce7QD8e+/btE6NHjxZ6vV7ExsaK22+/Xfz11191jqlOp1Pcd999Ii4uTkiS5PMcp7+ux/bt28XYsWOFXq8XWq1WjBw5UmzatMmnTaDvir9ppY0l/TdAakE7d+7EhRdeiM8++yyo7jUCli9fjquvvhobNmzAkCFDQh0OEVG7IusxAy3h9FKVHq+//joUCkVQXWtydOY+c7lcePPNN2EwGGS/IhoRUXOQ9ZiBlvDyyy9j27ZtGDlyJFQqlXdKzR133NGiI0Xbkvvuuw82mw0XX3wxamtr8dVXX2HTpk2YO3dui03pJCKSE94maGarV6/Gs88+i3379qGqqgopKSmYOnUqnnrqqVa7lGWoLVq0CK+++iqOHDmCmpoadO/eHXfddRfuvffeUIdGRNQuMRkgIiKSOY4ZICIikjkmA0RERDIX1E1rt9uN/Px8REZGNlmhHCIKnhACFosFiYmJzVpet6nx2EEUWsEeO4JKBvLz8znynagVyM3NRadOnUIdRtB47CBqHRo6dgSVDHiWBM3NzW3RWslEdEplZSWSk5PrLKXd2rW1eInaq4a+i0ElA57uPYPBwGSAKITaWld7W4uXqL1q6LvYdm4+EhERUbNgMkBERCRzTAaIiIhkjskAERGRzDEZICIikjkmA0RERDLHZICIiEjmmAwQERHJHJMBIiIimWMyQEREJHNMBoiIiGSOyQAREZHMMRkgIiKSOSYDREREMsdkgIiISOaYDBAREckckwEiIiKZYzJAREQkc0wGiIiIZI7JABERkcwxGSAiIpI5JgNEREQyx2SAiIhI5pgMEBERyRyTASIiIpljMkBERCRzTAaIiIhkjskAERGRzDEZICIikjkmA0RERDLHZICIiEjmmAwQERHJHJMBIiIimWMyQEREJHNMBoiIiGSOyQAREZHMMRkgIiKSOSYDREREMsdkgIiISOaYDBAREckckwEiIiKZYzJAREQkc0wGiIiIZI7JABERkcwxGSAiIpI5JgNEREQyx2SAiIhI5lShDoCahxACQohzeg5JkiBJUhNFRERErRWTgXZq+/bt+OOPP1BWVtbo57jiiivQt2/fJoyKiIhaIyYD7dS2bdswb948HDlypNHPER8fj+zsbLzzzjuIi4vDokWLmjBCIiJqLZgMtFGLFy/Gc889F/Dx8vJylJWVwW63N/o1nnnmGSiVSpSVlUGlUiE9PT1g244dO+KWW27B1KlTG/16REQUGkwGWjG3243Kykrs2rWrzmNbt27F/v37m/X1CwsLff5d3+sVFxdjy5YtSE1NBQB06dIFHTt2hFqtbtYYieRKoVAgJiYGJSUlAccHKRQKaDQa1NTUBGyj0WgghDinC4f6SJKE8PBw2Gw2KBQKuN3ugO0iIyNhsVjOebwTnT0mA62YxWLB559/jjVr1tR5LCsrKwQRBVZVVYVff/0VJ06cAABMmzYNo0aNYjJA1AiSJNV7QpQkCVFRUbDb7cjIyMDRo0dhs9nqtAsPD4fL5UJcXBzKy8vhcDj8trFarYiOjkZVVVXQSYFWq0VsbCyOHz8esI3RaETPnj3x119/QaPRwOl01mmjVCqh0+lgsVig0WhQVVXFZCAEmAy0Mtu2bfOeUEtLS/Huu+9iz549IY6qYTU1Ndi9ezd2794N4NRBwGKxwGAwoGPHjujatSvi4uJCHCVR+6BUKtGnTx8cO3YMJpMJSqWyThtJkhAfHw+z2QyNRgOFwv9M8piYGAghoNFoUF1dHXQMKpUKkZGR6NChAwCgqKioThu1Wo2YmBh07twZRqMRO3bsgMvlqhOnWq1GREQEdDrdOQ16psZjMhBiQgifjP7tt9/Gl19+CeDUbYKqqqpQhXZOFi9ejK+++gqSJGHo0KG4/vrrMWjQICQnJyMsLAxKpZLTFonOQlhYGCIjI6HT6aDVajFw4EAkJiZi1apVkCQJKpUKGo0GJpMJarUaYWFhSEtLQ0VFBY4ePYqoqChUVlYCAKKioqBWq6FSqdCrVy+YzWbk5uYiJiYGlZWVEELUm7xXVFSguroaVVVVGDx4MABg9+7dEELAarXC7XbDarWivLwcO3fuxA033IDIyEhUV1ejpqYGZrMZDocDdrsddrsdtbW1SEhIQExMDKxWK+x2O5xOJ9xuN1wuF2pra1tkH8sZk4FWYOPGjd5usZycHO8Xti2rra31foFXrlyJlStXIiUlBR9++CF69uyJ2NhYhIeHMyEgOkOgWwRdu3bFyJEjMXz4cMTFxeFf//oXAOCiiy4CAOzatQs9evTADTfcgM6dOyMyMhIbNmzweY5vvvkGCoUCN998M5KTkxEVFYXNmzf7tPnuu+/gdrsxa9asgDEuXboUmzZtwo033ujddu+990KhUGDnzp0oLy/Hli1bsHXrVvTr1w9HjhyBUqnEe++9h5qaGnz77bfIyspCTk4ODh8+jIyMDO97v/rqq1FRUYGioiLYbDaUlZXh0KFDjduZFDRJBHFzprKyEkajERUVFTAYDC0RV7t2+i53u90ICwsLOKimvVGpVLjqqquQnp6ORx99FHq9HgCYFDSgrX4HPXFT8AIlAxdccAH0ej22bNmCxMREZGdn12kTFxcHnU6H48ePw2g0ory8vE4bk8kEg8GAEydOIDo6GiUlJU0af//+/VFQUIC8vDyf7Z6xAVVVVbjmmmuQlZWFP//806eNJEmIjY1FeXk5TCYTHA4HKioqmjQ+uWro2MFkIAR++uknPPXUU95/n/mFaO88lQ0zMzPx2GOP4dJLL0V8fHyow2rV2up3kMnA2atv8KBOp4PRaERRUVGde+8eKpUKERERsFqtAdt4ZhlIkgSr1dpksddHqVTC5XJBr9ejtrbW72BGSZIQHR2NyspKv49T4zV07OBtghZSWlrq/WLm5ORg586doQ0ohDylkvfs2YNjx4556xd4BiRxBgLJWX3XZ9XV1bDZbPX2JDqdzgZH5Lvdbr+zD5qT5/hXXV0dMDYhBMrKyjibIASYDLSQGTNmeLv1zGaz3yk2cuNyuTBv3jwsWrQIarUaGRkZmDNnDtLS0kIdGlGrFcwtxdZ8Mm0ottYce3vGZKAZCSFw9OhRrF69Gps3b8bJkydDHVKrk5+fj/z8fACnqiYuXLgQvXv3xpQpUziOgNolpVKJhIQE7xTiMykUCkRHRyMsLAwFBQUBT44qlapdXVSEh4ejpqYm4PuSJAkpKSneKYy1tbVMHJoQlzBuRkIIZGVl4aOPPuIgmCAUFxdj+fLlWLJkSahDIWoWSqUS8fHx3nvn/hJepVKJlJQUOJ1O9OjRA2FhYQGfCzg15bAtJ84qlQpxcXEQQkCtVvuddqzVajF8+HA4nU4kJiYiPDw8YN0EahzuzWZSW1uLNWvW4LfffsOuXbs4TzYINpsNhw8fxo4dO/DTTz+hoqJCNrMsSB5UKhVSU1NRVlbmHcDnr02fPn2gUCgQHx/vt6CQp51KpYJarW7TyYBSqYTBYIBGo0FcXJzf96LRaHDeeechMjISo0aNCpggUePxNkETslqt3hrgRUVFGD9+PEfEniW3240TJ07g//7v//Dzzz8jPT0der0e4eHhUKn4caW2R5IkKJVKqFQq6PV66PV6xMTE+FTs02g00Gq1iIiIQHR0NAYOHIiamhqsWrUKSqXSOxJfkiQoFAooFApERERAqVS2+Vr+drsdhYWFSEtLg16vx8GDB2Gz2WC32+FyueB2u1FeXo7PPvsMjz32GAYMGIDdu3dDpVLBbDajpqYGQHBjKSgwHl2b0M8//4z169fDbrejuro64LQeCs4HH3yA2NhYDBgwAEOGDEHnzp3b9BUQyZNOp0NMTAw6deqExMRE/PzzzwDgnXJZWVmJzMxMXH755Rg1ahTS09Px4IMPQpIkXH755QCA3377DYWFhd6SvSqVCsXFxQDgnX3jqdjX1mi1WmRmZgI4dWt18uTJqK2txdGjR72rrx49ehQ6nQ5vvfUWDAYDlixZgpycHCxYsABbtmyBzWZrF8XaQol1BpqAZxc+8cQTeOutt86qvjc1zGQy4YknnsDtt9+OqKioUIcTEm31O8g6A6dK/7rdbrjdbqSkpGDv3r112vTt2xcKhQJHjhzBhRdeiLVr19Zp47lPbrPZoFarm22VwVDSaDRQq9WwWq3o3LkzbDYbCgoKfNoYDAakp6dj7969GD9+PGpqavD111+HKOK2g3UGWsDOnTtx11134c8//2RvQDMwm82oqqrivqU2qby8HLGxsTAYDAHL6u7YsQMmkwnx8fE4evSo3zae7vD23jtWVVUFrVaL3NzcgLdZ9+7di6SkJKxZswalpaUtHGH7xGTgHAgh8M0332DPnj0oLCzkyaoZ7dq1CytWrEBaWhouvvjiUIdDdFbKysrqrS8ihIDZbA7q/r8Qol32CgDwDrS22WwB94PFYgEAHDlyxNvjQueOycA5euedd3D48GEUFhaGOpR2bf369cjPz8eYMWOYDFCbE8xJSwjRruoGnIv6EiLPY9xXTYvJQCM4nU7k5+fjX//6F7Zs2YKKioo2PZq3LSgrK4PNZkN1dTXcbjeeeuopaLXadt9lSm2LJEmIi4sLWGDMM7WwtrY2YNEhaphCofCbXEmShIEDB2Lr1q0AwN7as8A6A2eptrYW+fn52L17N958802YzWYmAi3EZrNh3759+OCDD7z3T4laC4VCAZPJ5J32509YWBgyMzOhUCiQnJwcsIYA+ZIkyae2gL+LgLCwMJx//vlQqVRIS0vjEulnicnAWSotLcWPP/6INWvWhDoU2WLyRa1RWFgYkpKSUFpaGvBEpNFokJmZicrKSiQlJTEZCJInGfDUXPC3b9VqNTIzM5GdnY1rr70WOp0uBJG2XbxNECS32w2r1YodO3bgpZdeCjjil5qf2+1GSUkJlEoljEYjs38KKU8hII1G411iuKSkxPu4QqGASqWCRqNBhw4dkJGRgQEDBmDt2rUc/BYkt9uN6upqGAwGSJLkHWDodru9FwfV1dX44osvcO+992L8+PH49ttv4XA4UF1dDYfDwYuIBrDOQJDKysrw0Ucf4bXXXkNJSQkrC4aQQqFASkoKhg8fjnfeecfbJduek4K2+h2UQ50BvV6Pjh07wmQy4c8//6zzeHx8PDIzM72DX2+++WafxwsLC3nbqwGSJCE8PBzAqXEXOp0OkiTBarXCbrfD7XbDbrdDq9UCONVLs27dOuzcuRMLFizAn3/+CafT2eLLNrcmrDPQRFwuF8rLy+sUwKCW53a7kZ2djcTERBw6dAi9evWCRqMJdVgkQ0ajEU6nE0VFRQGv8jt27Ii8vDzMmzcPBw4c8C5lTsETQnhP5CqVyju90HPy90xJ9BR8czgceOWVV7Bu3TqkpaVhxIgR+O6770IQedvBMQPUZv3+++8YPHgwDh8+zC5AComKigqEh4cjMjIy4El+z549KC8vR1xcHPbv39+yAbZzVqs14NX+smXL0Lt3bxQUFOCHH35o4cjaHvYMBOGjjz7CH3/8gd9//z3UodBp3G43bDYb5syZg6effhr9+/cPdUgkQ2azGZIkBewZ8Ixx4TLmTSOY+gKeokzr16+Hw+HgFMMgsGegAU6nEzt37sS6detw+PDhUIdDfvz222945513sGHDhlCHQjLkcrkaPEG5XC7U1NRwbEALq66ubrfVGpsak4EAhBDYuXMnXnjhBfz66684ceKErAeftGZmsxn79u3jeA4KKYPBgPHjx4c6DNmLjo72TkMMJDY2FnFxcdBqtVAoeBoEmAz45XK5cOjQIXzwwQd46aWXsGvXLlit1lCHRfWoqqrCkSNHcOTIkVCHQjIVGRmJsWPHAjg1g6A9z25prVQqFdRqNaKjo6FWq+v8DRQKBRISEhAWFobU1FTodDomA//FveCHw+HAf/7zH8ybN4/dem1Efn4+vvnmGy5lSiGjUCig0+mg1+uRkpLCgkItzFOYqLi4GB07dvS7/5VKJbp27QqLxYIePXpwFtJpmAz44XK5sGHDBo5Qb0PKysrwxx9/YP369SgvL/dOMSJqbgqFAuHh4YiJiUHPnj1xwQUXYNu2bVxIp4UJIWC1WhEREQGNRgOVSlXnqt/hcGDTpk0YNGgQ+vfvD5PJBI1Gw94BcDaBD8/Jn0lA2/Xtt9+ia9eumDRpEubNmxfqcKidUyqV6NKlC6655hqMHj0aI0aM8G4HuLJeKFRXV2PHjh3o06cP8vLy6hSJU6lUWLduHbZv345HH30U2dnZWLduHY4ePSrrlSOZDJyhqKgI7733HguDEFGDevXqherqaqxZswZarVa2J5LWyGq1oqqqylsq2lOYyJMYOBwO/PLLL9izZw+0Wi0yMjKwc+fOEEYcWkwGTiOEQGFhIf75z396PzhERIEcOHAAQghER0eHOhQ6Q25uLoxGIywWS8AaDzk5OYiMjERZWRmysrJaOMLWhcnAaX755Rfcd999nEJIREHxFLM5cOAAp7a2Ig6HA06nEzU1NQGLQVVVVXl7gJ1Op+wLEzEZ+K+SkhLs378fBw4cCHUoRNTGWK1WTj9uRYQQ3lUNA3G73ewBPg2HUP7X4cOHsX379lCHQUStVEN1A3Q6HYYPH46wsLAWiogaq3fv3jAYDN4CRfWRS70IJgM4VcFu69atXHugnXA6ncjNzcXq1as5M4SaTEPTzzwVCJVKJUwmk2xOIm2JJEkwGAxITExEeno6EhMTG6w1oFAoZPG3ZDIAYMuWLfj555+59kA7UV1djZ9++gk33HBDvd2ERE1JqVQiKioKQgikpqay6FAr5KlAuG7dOlx22WXQ6/WhDqnVYDIA4N1338WKFStkP4CkvRBCwG63w2w2s2eAmkR9V4aSJEGlUsFoNKJXr17o3bs3du3axWmGrZDL5cLBgwdx3nnnoVu3btDpdEza/osDCImI6iFJEhQKhd+LhbCwMKSkpGDixIm47LLLMGbMGABAeHg4AKCmpoYJaSu0Z88evPjii3jwwQcRExODdevW4eTJk3XaKZVK2VwkMhkgImqklJQU2Gw2rFu3DgaDgdOS25hDhw5h165dkCQJKSkpOH78eKhDChneJiAiaqTs7GyUlZVBqVRyPYw26Pvvv0fnzp3RoUMH5OXlhTqckGLPABFRPYQQAbuKPcVq9u3bJ/uTSVtz/PhxuN1ulJSUoLa21u/fWC63CACZJwMbNmzA0aNHuQ4BETWaEAIWiwUWiyXUodBZ8NzSYeGhU2R9m2D9+vX46KOPZF+Tmojqp1AoGpyGptPpMHTo0BaKiM6FVquFSqWCVqsNOJtAo9FAq9UiMTEROp2uhSNsebJMBoQQKC8vx9atW7Ft2zaYzeZQh0RErZharW5wMSKTyYQbbrihhSKicxEREQG9Xo+oqCi/FQiVSiU0Gg1iYmLQs2dPREVFhSDKliXbZGDdunVYtWoVB/20c5WVlbDZbCw+RERepaWlMJlMAXsFdDodLBYLDAYD1Gp1C0cXGrJNBhYtWuRd15raJyEE/t//+3/47rvvAi5hStQUPLUImHS2D5WVlTAYDDAajbIpSiTbZGDZsmWsECYDr732GlatWoXKyspQh0JtlF6vR0xMTL1z0Pv27Yvbb78dDzzwQAtGRufC6XRCpVIFXHOioqICOp0ONpsNKpXKW0iqvZJlMkBERBSMqqoqVFRUQAgBrVYb6nCaDZMBIiKSpbKyMoSHh9d7krdYLFCpVFAqle26wqSs6wwQETXEZrM1OL7oyJEj+Oyzz1ooImoKJSUlcDgc9a4/sGPHDlitVm9xqfa8zgSTAWr35LAWOTUfl8vVYCW6iooKDlJtY2pqagDUX2WwpKSkpcIJOd4mICKqhyRJiIiIQHx8PFSqutdPCoUCSqUSSqVSFvPR5WbAgAFISEgIdRjNjskAtXvtuWuPmo9npLlKpUJUVBQyMjIQFhZWJyFQKpXe0eapqakAEHCEOrU9t956K/r06RPqMJqd7D6xbreb3XlE1CCj0QitVutNCPR6PWw2G4xGo8+tJ4fDAZfL5TP1TKPRhCJkokaTXTJgNpsxduzYUIdBRK1caWkpVCoVkpKSkJiYiJiYGMTExKC0tNRvb5NCoYBWq4VWq23Xo87lRJIkCCFk0bsouwGELpeLCxMRUVAcDgeio6PRpUsX/P7773UKz2g0Gu94gsTEROzfvx+SJHkXtmEp7Lbtyy+/xJNPPokDBw6EOpRmJ7tkgIgoGAaDAQ6HAyUlJYiKikJ+fj6qqqqQmJiIgoIC79VidXU1SkpKEB4eDqvVyvVO2hFPz4AcyOo2QWlpKZYvX84vKxE1qKqqCm63GwqFAlarFW63G263G8XFxd4ThN1uh8PhgNvthtPpRHV1dbuuUkftl6x6BqxWK/bt2we73R7qUKiFKBQKjuymRnG73XA4HCgqKkJFRYV3XvrpBYhO7x04ceIE3G43jy/tyKuvvor8/PxQh9EiZJUMuFwuWCwW2XT7EPDYY4/hoosugtFoDHUo1Aa53W5UV1c32JvocDi8SQIXQGs/NmzYEOoQWoyskgGAc87lRJIkPPfcc34LxRCdDc8Sxf6q1alUKu9PTEwMcnJyQhAhNZeMjAyUlJSguLi4XQ8GZf8pEVE9JEmCWq32u4StJEnQarWIiYlBamoqLr/88hBESM3p9ttvR//+/dv9RQWTASKiRoqIiIDdbkdNTQ3i4+NDHQ41A61WC5fL1e7HgjAZICJqJKvVCgAwmUwwGAwhjoao8dp3v4cfXMGOiM6GUqmEWq1GbW2t38dramogSRISEhKQlZUFk8kEs9ncskFSs/jqq68wZ84c7N+/P9ShNDv2DBARNQG3243y8nJUVFTAYDDwwqMdUKlUcLvdshh4LqtkQJKkdj8IhIhCw1OToEOHDrBarbI4gbRnOp3OW2xKDmSVDERHR+Paa6/F4MGDQx0KNSOlUom4uDj069ePV2d0zlwuF2pra/1OK/Q4efIkNm7ciJMnT6Kqqoq1BtoBu92O+fPno7CwMNShtAhZXSZHRkZi6NChmDlzJjZt2hTqcKiZaDQaZGRk4KabbmIyQOdMCNHgyb2yshKVlZUtFBG1BIfDgbVr14Y6jBYjq54BAAgPD8fkyZNDHQY1o7CwMKSlpWHGjBksRUzNTqlUIiwsDBEREUhISAh1OHQW4uPjERkZCaVSGbCNWq2u9/H2gkdKIqJGUqvV0Ov1iI6ORmpqKiZMmBDqkOgsXHbZZejZsyf0en3AC4eoqChotVoolcp23dPIZICIqJG6desGk8kEu93OokNtUHx8PGpqauB2uwOuX6JSqSCEgEqlatcrUjIZICJqpAMHDkCSJAwbNgx9+/YNdTh0ljp06IDExEQAQHl5ud82kZGRiIqKgkKhaHDBqrZMVgMIqf0bPnw45s6dy6s0ahFqtRpFRUU4evQozjvvPJhMJowePdq72l1tbS2nGLZCCQkJmDNnDl5++WXk5ub6nT5oMpnQqVMnZGVlyWKqKJMBaleio6PRv39/DhykFuFwOOB0OlFTUwOz2Yzdu3fjr7/+wvnnn4/t27e3+xNIW6NQKBAbG4uTJ0+irKzMe4vgTAaDAZWVld5ponL4OzIZoHZFqVRCo9GEOgySESEEhBCw2+0oKChAREQEDh06xFoDrZDb7UZFRQViYmJQVFQEh8Pht53VaoUkSbDb7Sw61J5JkoSbbroJPXr0gFqtDnU41ARiY2MxbNgwDBs2LNShkAyVlJRg69atKCkpQW1tLcrLy2VxNdkW1dbWoqamBuvWrQs4BsDTG1BRUVFvsan2RJY9AwqFAvfccw+sVisKCgoCZofUdnTo0AFjx47levIUEhUVFaioqAh1GBSkyspK7Nmzp942bre7XQ8YPJNsewYGDRqE+Ph4WRSTaO8yMjIwYMAADBw4EOedd16owyGZUalU0Gg0UKvVCAsL89tGoVBApVJ5/0utQ4cOHeqtMSAn/FRSm/f5558jMzOTX2hqcQqFAiaTCQaDAXa7HSqVCtnZ2T5tJElCREQEwsPD4XQ6ERERIZt6963d5MmTsXXrVuzevRtVVVWyuSXgD4+eRESN1LlzZ6jValgsFnTu3Nlvm7i4OISHh8NmsyEmJqZlA6R6eYoOmUwm9OjRI9ThhBSTASKiRjp27BgAoG/fvkhLS/Pb5uTJk3C5XOjatStSUlJaMjxqQL9+/TBs2DAIIXDgwIFQhxNSsk4GXn31VeTk5GD8+PGhDoUaYeTIkcjLy0Pv3r3bdc1wat2Kiorw+++/Y8eOHRg3bhwuuOACn8+jJEmwWCzIzs7GiRMnkJmZie7du0OpVPLWVguTJAmRkZEYO3Ys9u3bh2effRaffvopzGYzIiMj23W54YbI+pOo1WphNBoxfPhwDBw4ENHR0aEOiYIQFhaGCRMmYOTIkYiOjm73C4hQ6+Z2u+F0OlFVVYVDhw7h4MGD6Nmzp3dwshACLpcLDocDNpsNubm5OH78OBITE5kMtDClUomePXti06ZN2LFjB3Jzc1FWVoaqqipUV1fDZrOFOsSQ4QBCAH/729/gdDqxaNEilJWVhTocaoBWq8WDDz6ITp06sU4EtRoOhwNZWVmIjo5GUVFRnWI1nqSguLgYERERKCsrk/WAtVBwuVzIzs6GWq3G8ePHUVtbK5uiQg1hMoBTK4916dIFOp0u1KFQEFQqFc4//3xER0ezR4BaBbvdjtLSUjidToSHh/td9MblcsFiscDhcECtVstqDntrIYRASUkJDAYDfvzxR1RVVYU6pFaDycB/hYeHy/p+UVsgSRI0Gg1iY2OZBFCr4nA4YDabASDgCcblcnkTALvd3lKhkR+VlZVYt25dqMNoVXjD6r8SExORkZHBcQOtmEajQUpKCsaOHRuwuAtRa6RWq6HVaqFWq4NKZJnsnruwsDBoNJp6C8up1WpvMSi5Y8/Afw0YMABRUVHQarV4+eWXQx0O+REZGYnhw4fj9ddfD3UoREFTKpVISkpCUlISCgsLUVBQAKvVWu/veOoSUON1794dAFBYWBhwrYikpCS4XC5UVFSgqqpK1uMHmA6dplu3brjhhhtCHQYRtSMjR45Ep06dcPz4cQwZMsRvm4iICN6mbGLdu3dHWVkZhBCIiory2yY2NhY2mw3R0dFIT09v4QhbFyYD1CaYTCZ06tQJsbGxoQ6F6Kz8/PPPqKqqwr333oupU6f6bWOz2eB0OmEymWA0Gls4wvYpMTERKSkpcDqdAWeJDRo0CKNGjQKABhcuau94m+AMkZGRGDx4MDZt2hTqUOg0M2bMwMyZM9GxY8dQh0J0VjQaDY4fP4633noL3bt3x6ZNm7Bw4UJ8/PHHcLlcsFqtUCgUUCqVcDqd0Gg03hooFRUVEELAbrfLugs7GCqVCikpKUhMTMRVV12FL7/8EmazGXFxcYiJiUFtbS0KCgogSRKGDBmCgQMHYsmSJSguLubKtWAy4EOSJCQkJOCFF17Azp07YbVa8eKLL3IKUAjFxcVh0qRJ+Nvf/oZOnTohIiIi1CERnRW73Q6Hw4GqqipIkoQVK1ZgyZIlGDt2LFauXAmXywW32w2XywWFQgG3243KykoAp24fWK1WJgJB8NRx2LVrF3r16oXDhw/7zOwQQkClUuGCCy7Arl27EBkZifLyctTU1IQw6taDycBpJEmCVqvFRRddhO7du8NsNuO1115jMhBCBoMBo0aNQkZGBmcQUJvkOZELIWCxWLBmzRpERUVh9+7d3kGCQgjvADe32+29UpUkiYWJguR2u1FaWgqDwYADBw6gqqoKtbW1Pm0UCgXy8vKgUCiQm5vLHoHTMBk4g0KhgF6vh06nQ3R0NO666y4sX74cx44d4+jeFiBJEkwmE6ZNmwbg1Kpi559/Pu+jUpsnhIDVasW+ffsQGxuLgwcP1hnhLoSA0+n0/pv1CIInhEBVVRX0ej2OHTvmsx893G43CgoKYDAYUFhY6LeNXDEZCECSJOh0OrzwwgsoKCjAyZMnmQy0AKVSibi4OLz22muca03tihACtbW1qK2tRUlJScA2PEGdm8LCwgbbeG7D0P9wNgERUSugUCi8a20ESoQlSfIWyJFzsuwZcFnfPlCpVFzE7CwwGQiCWq1usJIVnRulUgmNRuP9IZITSZIQHx+PtLQ0hIeHB1wnJSIiwluOOzw8vIWjbD2MRiM6duwIvV4f8GTfqVMnxMfHIyIigglBEJgMBOGdd97B9u3b8dhjj4U6lHZJp9Nh9uzZsNlssFgs+Ouvv/jlJVm5+OKL0b17d2RnZ6N///5+26SmpiIlJQXl5eVISEho4Qhbl8TERJSXlyM9PR19+/b128ZTUCglJQUZGRktHGHbw2QgCJ5ZBoMHD8b1118Pk8kU6pDajREjRmDatGkYPHgwJEny/hDJyebNm1FSUoJbb701YBXU48ePo7S01DtHXs4mTpyIG264ASUlJdi+fbvfNsOGDcNll12Gmpoa7N27t4UjbHs4gDBIERERGDlyJJRKJbZv3+5doYzOzdixY3HVVVehc+fOoQ6FKGQSExOh1+uxc+dOFBUV4fvvv8fmzZuxbNkyOBwOZGVlISwsDHFxcTCbzRBC4L777kNBQQG2b98Op9OJgoKCdjswLiIiAn379sW4ceOQlpaGTz75BJWVlejYsSNiYmJQWVmJw4cPIykpCcOGDUO/fv2wZMkS71LSSUlJsNvtOHnyZIjfSevFZCBICoUCWq0WgwYNwsKFC70Ljdxyyy3Iz89nUZCzpNVqsXDhQgwYMADx8fGyvv9JVFpaCovFAqVSiePHj8NkMmHt2rXo06cP1q9fD6vVipqaGthsNu+xqLa2FidOnEBiYiKOHTvW4OJHbZlKpUJ8fDw++eQTXHDBBdi0aZNPDQGn0+mtQLhu3TrY7XYcOnTIZ5/wGF0/JgNnyWQyYfDgwQBOTQPSarXs1j4LkiQhJSUFs2bNwhVXXMHFWYgA1NTUoKamBgqFAi6XCz/88AMAYO/evSgtLfVON7Tb7ZAkCU6nE3v27IFSqURubi4qKira9ZTE2tpa7N27F8XFxdi0aROKi4vrnNxP3xcHDx6ExWJhUaGzwDED1GKUSiU6dOiASZMm4a677mIiQHQGTy2CEydOQKVSYd++fX4LDzmdTlRVVUGhUKCwsLDdn/TsdjsOHz4MSZJQWlrq9yrf5XIhJycHkiShqKiIlRvPEnsGzlFGRgaMRiNcLhfy8vL8Zqxy1qVLF+/yoRqNBunp6bj//vu9c6WJ6H88yQAAHDlyJGAbz8nfc09cDoQQAVcfPJ3FYoHFYmmBiNoXJgPn6P333/dmoHPnzsXHH3/svU/VnrvtGqJQKKBQKDBnzhyMHz8ewKlbBBqNBgaDIcTREbVtkiTVKWV8Os+CR+fyHKHimVHEi6qWxWTgHEiShJiYGO+/u3fvjvPPPx81NTWw2+3YtWtXq/yytYTExETExcUhNTUV8fHxoQ6HqN1Qq9UwGAwoKyvze3zRarWIiYlBbm4uwsPD/a7Kp1KpoNPpUFFR0RIhn5Xo6GiEhYXBbDYHLAGv1WpRU1PDhKEJsa+2Cd1zzz349ddf8ccff2DlypVQqeSba91333346aefMGrUqFCHQtQuGI1GxMTEQAgR8DZb586d0atXLxQWFgacrhsVFYVOnTqhsrIyZDVTevfujaSkJL+PJSQkoLy8HH379kXv3r3rfR6tVsuxR01EvmerZnD6rIKYmBj88MMP3sx1xowZyM3NDVVozWr06NGYPXu2z7YePXrAYDBwpgVRE6moqIBWq0WnTp1gNBr9LnaUk5MDq9WK0aNHIzExEZ9//nmdNmazGXa7Henp6YiMjMTvv//eEuEDAAYNGgSFQoEDBw4EHO8wYcIE5ObmYv369QGPmUajEREREX6XKabGYTLQTMLCwnDxxRd7/x0XF4fS0lLY7fZ2MZZAr9d7r05SU1MxZMgQn8fVarWse0aImlpCQgISExOhUqmgVqsxdepUlJaWorCw0DuSPiEhAQMGDEBERASMRiO+/vprHD9+HFu2bEFtbS22bNkCrVaLcePGISwsDAqFArNmzUJlZSVycnJgt9uxdu1aAMCdd94ZMJbvvvsOR44cwYMPPhiwzZo1a/Dbb79h+vTp6NOnDwBg//79EEJg5MiRcLlc2LNnDzZu3Ihx48bhsssuw3nnnYclS5ZAq9Vi6NChcDgcyMvLw++//47Y2FhERkZCq9V6EwmTyQQhBOx2OwvBnSNJBHFTu7KyEkajERUVFRz81Uhr167Fpk2bsGLFCmzZsiXU4Zyz9957Dx07dgRwakGQCy+8MMQRtW9t9TvoiZvOXUREhLdLXKlUIjk5GUVFRdBqtaioqEB5ebl3vIBnIaMLL7wQx44dQ3x8PI4dO4asrCyo1WpvF72n7kd5eTkiIyNRVlaGY8eOAQB69uwZMJYTJ06gsrIS6enpAdsUFBTAYrFgxIgR3l6M0tJSnzbl5eWwWq0YM2YM8vLy0KdPH6xatcrngslqtaK0tBRRUVGw2+3esQ6njxdwu91+p2DS/zR07OClWwsZOXIkYmNjoVAokJycXOfxw4cP49ChQ34H+4RSZGQkUlNT6xwYJkyY4E0GiKj52Ww274A6hUIBIQRKSkqg1+tRXV3tPSGazWbvzJ28vDwAQF5eHgoLC73Hl9NPygcPHgRwasEwi8XiPakWFxc3GNOGDRvqfVyj0WDTpk3eOPyJiIjAjh07kJeXh+PHj+P48eN1BgZKkoSqqirU1NTA5XLBZrPJdnB2c2Ey0IKSk5Nx5ZVXYujQoXUeW7p0KfLz81tdMhAVFYUhQ4ZgypQpPtt5tUcUOkII77HC35x6IQRcLhfKy8sRHx/v9wTrUVNTA61WG3B2wrmora2tNxEATiU5Bw8ehMlkClhQ6PT363Q6mQg0AyYDLchkMgUcvZudnY29e/fi5MmTqK6uRmlpaUgWHUlPT/cZqdylSxdcdNFFGDZsWIvHQkT+CSEaLKzjKUwUzOI8oV7XQAgRdAElDhhsHhwz0AqtX78e8+fPx5dfftnir11SUsL1Flqhtvod5JgBotaBYwbaoNTUVPztb39Dr169Wvy11Wp1i78mERGFFpOBVig5ORnJycmYOHFii782ewSIiOSHyUAr5Dkh88RMREQtgeWIiYiIZI7JABERkcwxGSAiIpI5JgNEREQyx2SAiIhI5pgMEBERyRyTASIiIpljMkBERCRzTAaIiIhkjskAERGRzDEZICIikjkmA0RERDLHZICIiEjmmAwQERHJHJMBIiIimWMyQEREJHNMBoiIiGSOyQAREZHMMRkgIiKSOSYDREREMsdkgIiISOaYDBAREckckwEiIiKZYzJAREQkc0wGiIiIZI7JABERkcwxGSAiIpI5JgNEREQyx2SAiIhI5pgMEBERyRyTASIiIpljMkBERCRzTAaIiIhkjskAERGRzDEZICIikjkmA0RERDLHZICIiEjmmAwQERHJHJMBIiIimWMyQEREJHNMBoiIiGSOyQAREZHMMRkgIiKSOSYDREREMsdkgIiISOaYDBAREckckwEiIiKZYzJAREQkc0wGiIiIZE4VTCMhBACgsrKyWYMhIv883z3Pd7GtaGvxErVXDX0Xg0oGLBYLACA5OfncIyKiRrNYLDAajaEOI2ieYwcRhVZDxw5JBJG6u91u5OfnIzIyEpIkNWmARNQwIQQsFgsSExOhULSdu3s8dhCFVrDHjqCSASIiImq/2s4lBhERETULJgNEREQyx2SAiIhI5pgMEBERyRyTASIiIpljMkBERCRzTAaIiIhk7v8Dl9lv0LdEkOwAAAAASUVORK5CYII=",
      "text/plain": [
       "<Figure size 640x480 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "SVM HoG Prediction: [0]\n",
      "Random Forest HoG Prediction: [0]\n",
      "SVM HoG Accuracy: 1.0\n",
      "Random Forest HoG Accuracy: 0.9966666666666667\n"
     ]
    }
   ],
   "source": [
    "import os\n",
    "import cv2\n",
    "import numpy as np\n",
    "import random\n",
    "from keras.utils import to_categorical  # Updated import\n",
    "from skimage.feature import hog\n",
    "from sklearn.model_selection import train_test_split\n",
    "from sklearn.svm import SVC\n",
    "from sklearn.ensemble import RandomForestClassifier\n",
    "from sklearn.metrics import accuracy_score\n",
    "import matplotlib.pyplot as plt\n",
    "\n",
    "dataset_path = \"/kaggle/input/four-shapes/shapes\"\n",
    "CATEGORIES = [\"circle\", \"square\", \"star\"]\n",
    "IMG_SIZE = 200\n",
    "MAX_IMAGES_PER_CLASS = 500  # Limit the number of images per class\n",
    "\n",
    "training = []\n",
    "\n",
    "def TrainingData():\n",
    "    for i, category in enumerate(CATEGORIES):\n",
    "        path = os.path.join(dataset_path, category)\n",
    "        print(path)\n",
    "        class_num = CATEGORIES.index(category)\n",
    "        \n",
    "        image_count = 0  # Keep track of the number of images collected for this class\n",
    "        for img in os.listdir(path):\n",
    "            if image_count >= MAX_IMAGES_PER_CLASS:\n",
    "                break  # Stop collecting images once the limit is reached\n",
    "            \n",
    "            img_array = cv2.imread(os.path.join(path, img))\n",
    "            new_array = cv2.resize(img_array, (IMG_SIZE, IMG_SIZE))\n",
    "            training.append([new_array, class_num])\n",
    "            image_count += 1\n",
    "\n",
    "# Call the TrainingData() function to process the dataset\n",
    "TrainingData()\n",
    "\n",
    "random.shuffle(training)\n",
    "\n",
    "X = []\n",
    "y = []\n",
    "for features, label in training:\n",
    "    X.append(features)\n",
    "    y.append(label)\n",
    "X = np.array(X).reshape(-1, IMG_SIZE, IMG_SIZE, 3)\n",
    "\n",
    "Y = to_categorical(y, 3)  # Use to_categorical from Keras\n",
    "\n",
    "def extract_hog_features(image):\n",
    "    # Convert the image to grayscale\n",
    "    gray_image = cv2.cvtColor(image, cv2.COLOR_BGR2GRAY)\n",
    "    \n",
    "    # Extract HoG features\n",
    "    features, hog_image = hog(gray_image, pixels_per_cell=(8, 8), cells_per_block=(1, 1), block_norm='L2-Hys', transform_sqrt=True, visualize=True)\n",
    "    \n",
    "    # Resize HoG features to the same size as the input image (200x200)\n",
    "    hog_resized = cv2.resize(hog_image, (IMG_SIZE, IMG_SIZE), interpolation=cv2.INTER_LINEAR)\n",
    "    \n",
    "    return hog_resized\n",
    "\n",
    "X_hog = []\n",
    "for image in X:\n",
    "    temp = extract_hog_features(image)\n",
    "    X_hog.append(temp)\n",
    "\n",
    "# Now, X_hog contains HoG features\n",
    "\n",
    "# Split the HoG feature data into training and testing sets\n",
    "X_train_hog, X_test_hog, y_train_hog, y_test_hog = train_test_split(X_hog, y, test_size=0.2, random_state=3)\n",
    "\n",
    "X_train_hog = np.array(X_train_hog)\n",
    "X_test_hog = np.array(X_test_hog)\n",
    "y_train_hog = np.array(y_train_hog)\n",
    "y_test_hog = np.array(y_test_hog)\n",
    "\n",
    "# Visualize the HoG features of an example image\n",
    "plt.subplot(121), plt.imshow(cv2.cvtColor(X[0], cv2.COLOR_BGR2RGB))\n",
    "plt.title('Original Image'), plt.xticks([]), plt.yticks([])\n",
    "plt.subplot(122), plt.imshow(X_test_hog[0], cmap='gray')\n",
    "plt.title('HoG Feature Visualization'), plt.xticks([]), plt.yticks([])\n",
    "plt.show()\n",
    "\n",
    "# Flatten the HoG features\n",
    "X_train_hog = X_train_hog.reshape((X_train_hog.shape[0], IMG_SIZE * IMG_SIZE))\n",
    "X_test_hog = X_test_hog.reshape((X_test_hog.shape[0], IMG_SIZE * IMG_SIZE))\n",
    "\n",
    "# Create SVM and Random Forest classifiers\n",
    "svm_classifier_hog = SVC(kernel='linear', C=1.0)\n",
    "rf_classifier_hog = RandomForestClassifier(n_estimators=100, random_state=0)\n",
    "\n",
    "# Train the classifiers\n",
    "svm_classifier_hog.fit(X_train_hog, y_train_hog)\n",
    "rf_classifier_hog.fit(X_train_hog, y_train_hog)\n",
    "\n",
    "# Predict the class using SVM for HoG features\n",
    "svm_prediction_hog = svm_classifier_hog.predict([X_test_hog[0]])\n",
    "\n",
    "# Predict the class using Random Forest for HoG features\n",
    "rf_prediction_hog = rf_classifier_hog.predict([X_test_hog[0]])\n",
    "\n",
    "print(\"SVM HoG Prediction:\", svm_prediction_hog)\n",
    "print(\"Random Forest HoG Prediction:\", rf_prediction_hog)\n",
    "\n",
    "# Evaluate the accuracies for HoG features\n",
    "svm_predictions_hog = svm_classifier_hog.predict(X_test_hog)\n",
    "rf_predictions_hog = rf_classifier_hog.predict(X_test_hog)\n",
    "\n",
    "svm_accuracy_hog = accuracy_score(y_test_hog, svm_predictions_hog)\n",
    "rf_accuracy_hog = accuracy_score(y_test_hog, rf_predictions_hog)\n",
    "\n",
    "print(\"SVM HoG Accuracy:\", svm_accuracy_hog)\n",
    "print(\"Random Forest HoG Accuracy:\", rf_accuracy_hog)\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "9f3d733f",
   "metadata": {
    "papermill": {
     "duration": 0.005012,
     "end_time": "2023-11-09T22:32:44.234237",
     "exception": false,
     "start_time": "2023-11-09T22:32:44.229225",
     "status": "completed"
    },
    "tags": []
   },
   "source": [
    "# > **> SVM HoG Accuracy: 0.9966666666666667\n",
    "# > > Random Forest HoG Accuracy: 0.9933333333333333"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "6937f283",
   "metadata": {
    "papermill": {
     "duration": 0.004929,
     "end_time": "2023-11-09T22:32:44.244545",
     "exception": false,
     "start_time": "2023-11-09T22:32:44.239616",
     "status": "completed"
    },
    "tags": []
   },
   "source": [
    "# LBP - SVM & RF "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "d33512e7",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2023-11-09T22:32:44.258125Z",
     "iopub.status.busy": "2023-11-09T22:32:44.257101Z",
     "iopub.status.idle": "2023-11-09T22:33:21.498516Z",
     "shell.execute_reply": "2023-11-09T22:33:21.497230Z"
    },
    "papermill": {
     "duration": 37.25199,
     "end_time": "2023-11-09T22:33:21.501845",
     "exception": false,
     "start_time": "2023-11-09T22:32:44.249855",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "/kaggle/input/four-shapes/shapes/circle\n",
      "/kaggle/input/four-shapes/shapes/square\n",
      "/kaggle/input/four-shapes/shapes/star\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "SVM LBP Prediction: [2]\n",
      "Random Forest LBP Prediction: [2]\n",
      "SVM LBP Accuracy: 0.9966666666666667\n",
      "Random Forest LBP Accuracy: 0.9933333333333333\n"
     ]
    }
   ],
   "source": [
    "import os\n",
    "import cv2\n",
    "import numpy as np\n",
    "import random\n",
    "from skimage.feature import local_binary_pattern\n",
    "from sklearn.model_selection import train_test_split\n",
    "from sklearn.svm import SVC\n",
    "from sklearn.ensemble import RandomForestClassifier\n",
    "from sklearn.metrics import accuracy_score\n",
    "import matplotlib.pyplot as plt\n",
    "\n",
    "dataset_path = \"/kaggle/input/four-shapes/shapes\"\n",
    "CATEGORIES = [\"circle\", \"square\", \"star\"]\n",
    "IMG_SIZE = 200\n",
    "MAX_IMAGES_PER_CLASS = 500  # Limit the number of images per class\n",
    "\n",
    "training = []\n",
    "\n",
    "def TrainingData():\n",
    "    for i, category in enumerate(CATEGORIES):\n",
    "        path = os.path.join(dataset_path, category)\n",
    "        print(path)\n",
    "        class_num = CATEGORIES.index(category)\n",
    "        \n",
    "        image_count = 0  # Keep track of the number of images collected for this class\n",
    "        for img in os.listdir(path):\n",
    "            if image_count >= MAX_IMAGES_PER_CLASS:\n",
    "                break  # Stop collecting images once the limit is reached\n",
    "            \n",
    "            img_array = cv2.imread(os.path.join(path, img))\n",
    "            new_array = cv2.resize(img_array, (IMG_SIZE, IMG_SIZE))\n",
    "            training.append([new_array, class_num])\n",
    "            image_count += 1\n",
    "\n",
    "# Call the TrainingData() function to process the dataset\n",
    "TrainingData()\n",
    "\n",
    "random.shuffle(training)\n",
    "\n",
    "X = []\n",
    "y = []\n",
    "for features, label in training:\n",
    "    X.append(features)\n",
    "    y.append(label)\n",
    "X = np.array(X).reshape(-1, IMG_SIZE, IMG_SIZE, 3)\n",
    "\n",
    "def extract_lbp_features(image):\n",
    "    # Convert the image to grayscale\n",
    "    gray_image = cv2.cvtColor(image, cv2.COLOR_BGR2GRAY)\n",
    "    \n",
    "    # Extract LBP features\n",
    "    lbp_image = local_binary_pattern(gray_image, P=8, R=1, method='uniform')\n",
    "    \n",
    "    # Resize LBP features to the same size as the input image (200x200)\n",
    "    lbp_resized = cv2.resize(lbp_image, (IMG_SIZE, IMG_SIZE), interpolation=cv2.INTER_LINEAR)\n",
    "    \n",
    "    return lbp_resized\n",
    "\n",
    "X_lbp = []\n",
    "for image in X:\n",
    "    temp = extract_lbp_features(image)\n",
    "    X_lbp.append(temp)\n",
    "\n",
    "# Now, X_lbp contains LBP features\n",
    "\n",
    "# Split the LBP feature data into training and testing sets\n",
    "X_train_lbp, X_test_lbp, y_train_lbp, y_test_lbp = train_test_split(X_lbp, y, test_size=0.2, random_state=3)\n",
    "\n",
    "X_train_lbp = np.array(X_train_lbp)\n",
    "X_test_lbp = np.array(X_test_lbp)\n",
    "y_train_lbp = np.array(y_train_lbp)\n",
    "y_test_lbp = np.array(y_test_lbp)\n",
    "\n",
    "# Visualize the LBP features of an example image\n",
    "plt.subplot(121), plt.imshow(cv2.cvtColor(X[0], cv2.COLOR_BGR2RGB))\n",
    "plt.title('Original Image'), plt.xticks([]), plt.yticks([])\n",
    "plt.subplot(122), plt.imshow(X_test_lbp[0], cmap='gray')\n",
    "plt.title('LBP Feature Visualization'), plt.xticks([]), plt.yticks([])\n",
    "plt.show()\n",
    "\n",
    "# Flatten the LBP features\n",
    "X_train_lbp = X_train_lbp.reshape((X_train_lbp.shape[0], IMG_SIZE * IMG_SIZE))\n",
    "X_test_lbp = X_test_lbp.reshape((X_test_lbp.shape[0], IMG_SIZE * IMG_SIZE))\n",
    "\n",
    "# Create SVM and Random Forest classifiers\n",
    "svm_classifier_lbp = SVC(kernel='linear', C=1.0)\n",
    "rf_classifier_lbp = RandomForestClassifier(n_estimators=100, random_state=0)\n",
    "\n",
    "# Train the classifiers\n",
    "svm_classifier_lbp.fit(X_train_lbp, y_train_lbp)\n",
    "rf_classifier_lbp.fit(X_train_lbp, y_train_lbp)\n",
    "\n",
    "# Predict the class using SVM for LBP features\n",
    "svm_prediction_lbp = svm_classifier_lbp.predict([X_test_lbp[0]])\n",
    "\n",
    "# Predict the class using Random Forest for LBP features\n",
    "rf_prediction_lbp = rf_classifier_lbp.predict([X_test_lbp[0]])\n",
    "\n",
    "print(\"SVM LBP Prediction:\", svm_prediction_lbp)\n",
    "print(\"Random Forest LBP Prediction:\", rf_prediction_lbp)\n",
    "\n",
    "# Evaluate the accuracies for LBP features\n",
    "svm_predictions_lbp = svm_classifier_lbp.predict(X_test_lbp)\n",
    "rf_predictions_lbp = rf_classifier_lbp.predict(X_test_lbp)\n",
    "\n",
    "svm_accuracy_lbp = accuracy_score(y_test_lbp, svm_predictions_lbp)\n",
    "rf_accuracy_lbp = accuracy_score(y_test_lbp, rf_predictions_lbp)\n",
    "\n",
    "print(\"SVM LBP Accuracy:\", svm_accuracy_lbp)\n",
    "print(\"Random Forest LBP Accuracy:\", rf_accuracy_lbp)\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "bc9e2ae2",
   "metadata": {
    "papermill": {
     "duration": 0.006851,
     "end_time": "2023-11-09T22:33:21.516133",
     "exception": false,
     "start_time": "2023-11-09T22:33:21.509282",
     "status": "completed"
    },
    "tags": []
   },
   "source": [
    "**SVM LBP Accuracy: 1.0\n",
    "Random Forest LBP Accuracy: 1.0******"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "8da79972",
   "metadata": {
    "papermill": {
     "duration": 0.006716,
     "end_time": "2023-11-09T22:33:21.529935",
     "exception": false,
     "start_time": "2023-11-09T22:33:21.523219",
     "status": "completed"
    },
    "tags": []
   },
   "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": "3.10.12"
  },
  "papermill": {
   "default_parameters": {},
   "duration": 339.506845,
   "end_time": "2023-11-09T22:33:23.168275",
   "environment_variables": {},
   "exception": null,
   "input_path": "__notebook__.ipynb",
   "output_path": "__notebook__.ipynb",
   "parameters": {},
   "start_time": "2023-11-09T22:27:43.661430",
   "version": "2.4.0"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}