problem_set_6.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {
    "deletable": true,
    "editable": true
   },
   "source": [
    "# <center>Digital Image Processing - Problem Set 6</center>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "deletable": true,
    "editable": true
   },
   "source": [
    "### Student Names:\n",
    "Luis Díaz,\n",
    "Juan Gómez,\n",
    "Jessied Marriaga"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "deletable": true,
    "editable": true
   },
   "source": [
    "### Problem 1"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "deletable": true,
    "editable": true
   },
   "source": [
    "Write a function that performs object recognition using $k$-Nearest Neighbors ($k$-NN). You can (and are encouraged to) use the functions\n",
    "you have written for previous exercises. Your function should run the following steps:\n",
    "\n",
    "1. Binarize the images in the training set. Use an appropriate threshold when binarizing each image. You can use an adaptive thresholding technique. Perform binarization such that pixels on the object of interest are equal to 1 and background pixels are equal to 0.\n",
    "2. Describe each object in the training set using Hu moments. This results in one feature vector per object.\n",
    "3. Repeat the two previous steps for the all images in the training set.\n",
    "4. The $k$-NN classifier predicts a category for each testing example according to the following steps:\n",
    "    1. Pick an image from the testing set.\n",
    "    2. Compute the distance from the feature vector associated to the testing example to all the descriptors from the training set.\n",
    "    3. Select the closest $k$ training examples (that is, the $k$ training examples with shortest distance to our testing image).\n",
    "    4. Find the object category associated to each of the $k$ training examples. This information is given to you in the training set.\n",
    "    5. Assign a label to the testing image. The assigned label is equal to the most common label among the $k$ closes training examples.\n",
    "    \n",
    "5. Repeat this process for all testing examples. Compute the accuracy of the classifier in terms of the number of correctly classified testing examples over the total number of examples in the testing set.\n",
    "\n",
    "You should test your functions with the three object classes <tt> key, chip, pen</tt>. Compare the difference in performance between a 1-NN classifier, a 3-NN classifier and a $k$-NN classifier. Summarize the performance of each classifier using confusion matrices."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {
    "collapsed": false,
    "deletable": true,
    "editable": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Performing image recognition on file: chip_00.jpg\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAXoAAADHCAYAAAAXg5iPAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fe030e59090>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Using 1 neighbours: \n",
      "Result:  Chip\n",
      "Neighbours:  ['chip']\n",
      "Distance:  [[  4.30672970e-10]]\n",
      "\n",
      "\n",
      "Using 3 neighbours: \n",
      "Result:  Chip\n",
      "Neighbours:  ['chip', 'chip', 'chip']\n",
      "Distance:  [[  4.30672970e-10   2.76392487e-09   3.48760243e-09]]\n",
      "\n",
      "\n",
      "Using 9 neighbours: \n",
      "Result:  Chip\n",
      "Neighbours:  ['chip', 'chip', 'chip', 'chip', 'chip', 'key', 'chip', 'chip', 'key']\n",
      "Distance:  [[  4.30672970e-10   2.76392487e-09   3.48760243e-09   1.42274885e-08\n",
      "    1.77863200e-08   3.96538447e-08   4.18112407e-08   1.51612383e-07\n",
      "    1.74008164e-07]]\n",
      "\n",
      "\n",
      "Performing image recognition on file: chip_01.jpg\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAXoAAADHCAYAAAAXg5iPAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nXwAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fe0252e7250>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Using 1 neighbours: \n",
      "Result:  Chip\n",
      "Neighbours:  ['chip']\n",
      "Distance:  [[  4.22728214e-10]]\n",
      "\n",
      "\n",
      "Using 3 neighbours: \n",
      "Result:  Chip\n",
      "Neighbours:  ['chip', 'key', 'chip']\n",
      "Distance:  [[  4.22728214e-10   6.71093514e-10   8.40389802e-09]]\n",
      "\n",
      "\n",
      "Using 9 neighbours: \n",
      "Result:  Chip\n",
      "Neighbours:  ['chip', 'key', 'chip', 'chip', 'chip', 'chip', 'chip', 'key', 'chip']\n",
      "Distance:  [[  4.22728214e-10   6.71093514e-10   8.40389802e-09   1.11849916e-08\n",
      "    2.70063918e-08   2.75501097e-08   2.97442284e-08   3.69042219e-08\n",
      "    4.17325623e-08]]\n",
      "\n",
      "\n",
      "Performing image recognition on file: chip_02.jpg\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAXoAAADHCAYAAAAXg5iPAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nNxgMhirn/wcj0U/AWetCDwAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fe024999f10>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Using 1 neighbours: \n",
      "Result:  Chip\n",
      "Neighbours:  ['chip']\n",
      "Distance:  [[  6.07116668e-09]]\n",
      "\n",
      "\n",
      "Using 3 neighbours: \n",
      "Result:  Chip\n",
      "Neighbours:  ['chip', 'key', 'chip']\n",
      "Distance:  [[  6.07116668e-09   1.11695861e-08   1.14444392e-08]]\n",
      "\n",
      "\n",
      "Using 9 neighbours: \n",
      "Result:  Chip\n",
      "Neighbours:  ['chip', 'key', 'chip', 'key', 'chip', 'chip', 'chip', 'key', 'chip']\n",
      "Distance:  [[  6.07116668e-09   1.11695861e-08   1.14444392e-08   1.26166810e-08\n",
      "    3.11944106e-08   3.17165068e-08   3.69322350e-08   4.39138503e-08\n",
      "    6.37061532e-08]]\n",
      "\n",
      "\n",
      "Performing image recognition on file: chip_03.jpg\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAXoAAADHCAYAAAAXg5iPAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nnP8H5JQwTCqMjasAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fe024c51f10>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Using 1 neighbours: \n",
      "Result:  Key\n",
      "Neighbours:  ['key']\n",
      "Distance:  [[  4.76545525e-10]]\n",
      "\n",
      "\n",
      "Using 3 neighbours: \n",
      "Result:  Chip\n",
      "Neighbours:  ['key', 'chip', 'chip']\n",
      "Distance:  [[  4.76545525e-10   7.38466566e-10   1.93053928e-09]]\n",
      "\n",
      "\n",
      "Using 9 neighbours: \n",
      "Result:  Chip\n",
      "Neighbours:  ['key', 'chip', 'chip', 'chip', 'chip', 'chip', 'chip', 'chip', 'key']\n",
      "Distance:  [[  4.76545525e-10   7.38466566e-10   1.93053928e-09   3.36680728e-09\n",
      "    1.39824472e-08   1.55576476e-08   2.45080898e-08   4.49741577e-08\n",
      "    5.75230636e-08]]\n",
      "\n",
      "\n",
      "Performing image recognition on file: key_00.jpg\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAXoAAACUCAYAAACdmeLWAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fe02529ef10>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Using 1 neighbours: \n",
      "Result:  Pen\n",
      "Neighbours:  ['pen']\n",
      "Distance:  [[  1.47171664e-09]]\n",
      "\n",
      "\n",
      "Using 3 neighbours: \n",
      "Result:  Key\n",
      "Neighbours:  ['pen', 'key', 'key']\n",
      "Distance:  [[  1.47171664e-09   3.01268770e-08   6.24026484e-08]]\n",
      "\n",
      "\n",
      "Using 9 neighbours: \n",
      "Result:  Key\n",
      "Neighbours:  ['pen', 'key', 'key', 'chip', 'key', 'key', 'key', 'chip', 'key']\n",
      "Distance:  [[  1.47171664e-09   3.01268770e-08   6.24026484e-08   1.18263323e-07\n",
      "    1.35525553e-07   2.04388115e-07   2.96533585e-07   3.33490448e-07\n",
      "    4.20447407e-07]]\n",
      "\n",
      "\n",
      "Performing image recognition on file: key_01.jpg\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAVoAAAEICAYAAAAeFzyKAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nmTAqtIqiKBNGhVZRFGXCqNAqiqJMmP8fzu6JuX1C2WIAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fe0247128d0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Using 1 neighbours: \n",
      "Result:  Chip\n",
      "Neighbours:  ['chip']\n",
      "Distance:  [[  1.59703903e-10]]\n",
      "\n",
      "\n",
      "Using 3 neighbours: \n",
      "Result:  Key\n",
      "Neighbours:  ['chip', 'key', 'key']\n",
      "Distance:  [[  1.59703903e-10   2.07687179e-09   3.39859851e-09]]\n",
      "\n",
      "\n",
      "Using 9 neighbours: \n",
      "Result:  Key\n",
      "Neighbours:  ['chip', 'key', 'key', 'chip', 'key', 'key', 'chip', 'chip', 'key']\n",
      "Distance:  [[  1.59703903e-10   2.07687179e-09   3.39859851e-09   7.40734318e-09\n",
      "    1.90519618e-08   4.92768422e-08   6.06278974e-08   7.34205656e-08\n",
      "    7.63458701e-08]]\n",
      "\n",
      "\n",
      "Performing image recognition on file: key_02.jpg\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAXoAAADFCAYAAABaSzmEAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fe025173650>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Using 1 neighbours: \n",
      "Result:  Pen\n",
      "Neighbours:  ['pen']\n",
      "Distance:  [[  8.34067748e-09]]\n",
      "\n",
      "\n",
      "Using 3 neighbours: \n",
      "Result:  Key\n",
      "Neighbours:  ['pen', 'key', 'key']\n",
      "Distance:  [[  8.34067748e-09   5.13180787e-08   9.16691576e-08]]\n",
      "\n",
      "\n",
      "Using 9 neighbours: \n",
      "Result:  Key\n",
      "Neighbours:  ['pen', 'key', 'key', 'chip', 'key', 'key', 'key', 'chip', 'key']\n",
      "Distance:  [[  8.34067748e-09   5.13180787e-08   9.16691576e-08   1.57496657e-07\n",
      "    1.77326854e-07   2.55082625e-07   3.57021861e-07   3.97467630e-07\n",
      "    4.91938465e-07]]\n",
      "\n",
      "\n",
      "Performing image recognition on file: key_03.jpg\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYAAAACcCAYAAAB2pHzrAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fe0249ece10>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Using 1 neighbours: \n",
      "Result:  Chip\n",
      "Neighbours:  ['chip']\n",
      "Distance:  [[  2.11204687e-12]]\n",
      "\n",
      "\n",
      "Using 3 neighbours: \n",
      "Result:  Key\n",
      "Neighbours:  ['chip', 'key', 'key']\n",
      "Distance:  [[  2.11204687e-12   6.60391353e-10   8.58153815e-09]]\n",
      "\n",
      "\n",
      "Using 9 neighbours: \n",
      "Result:  Key\n",
      "Neighbours:  ['chip', 'key', 'key', 'key', 'key', 'key', 'chip', 'pen', 'key']\n",
      "Distance:  [[  2.11204687e-12   6.60391353e-10   8.58153815e-09   1.20236106e-08\n",
      "    2.85172010e-08   4.08476062e-08   5.52463142e-08   9.24641625e-08\n",
      "    9.36225817e-08]]\n",
      "\n",
      "\n",
      "Performing image recognition on file: pen_00.jpg\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAXoAAADHCAYAAAAXg5iPAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fe02479dc50>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Using 1 neighbours: \n",
      "Result:  Pen\n",
      "Neighbours:  ['pen']\n",
      "Distance:  [[  2.10747531e-09]]\n",
      "\n",
      "\n",
      "Using 3 neighbours: \n",
      "Result:  Pen\n",
      "Neighbours:  ['pen', 'pen', 'pen']\n",
      "Distance:  [[  2.10747531e-09   2.94894775e-09   3.64686192e-09]]\n",
      "\n",
      "\n",
      "Using 9 neighbours: \n",
      "Result:  Pen\n",
      "Neighbours:  ['pen', 'pen', 'pen', 'pen', 'pen', 'key', 'key', 'chip', 'key']\n",
      "Distance:  [[  2.10747531e-09   2.94894775e-09   3.64686192e-09   5.21998160e-08\n",
      "    2.18390528e-06   2.60181332e-06   2.85355873e-06   3.18028879e-06\n",
      "    3.26734676e-06]]\n",
      "\n",
      "\n",
      "Performing image recognition on file: pen_01.jpg\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAXoAAADBCAYAAADB2nuSAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fe0248859d0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Using 1 neighbours: \n",
      "Result:  Pen\n",
      "Neighbours:  ['pen']\n",
      "Distance:  [[  5.70405155e-06]]\n",
      "\n",
      "\n",
      "Using 3 neighbours: \n",
      "Result:  Pen\n",
      "Neighbours:  ['pen', 'pen', 'pen']\n",
      "Distance:  [[  5.70405155e-06   6.60939304e-06   7.13467944e-06]]\n",
      "\n",
      "\n",
      "Using 9 neighbours: \n",
      "Result:  Pen\n",
      "Neighbours:  ['pen', 'pen', 'pen', 'pen', 'pen', 'key', 'key', 'chip', 'key']\n",
      "Distance:  [[  5.70405155e-06   6.60939304e-06   7.13467944e-06   7.16720024e-06\n",
      "    1.67654416e-05   1.78909577e-05   1.85416720e-05   1.93608248e-05\n",
      "    1.95747598e-05]]\n",
      "\n",
      "\n",
      "--------------------------------------------\n",
      "Acuracy and confusion matrix for K=1\n",
      "Accuracy:  0.5\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAUwAAAEmCAYAAAAJAaljAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nk5GTRzH6ugAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fe025079850>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Acuracy and confusion matrix for K=3\n",
      "Accuracy:  1.0\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAUwAAAEmCAYAAAAJAaljAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nlcekzQAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fe024e2b250>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Acuracy and confusion matrix for K=9\n",
      "Accuracy:  1.0\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAUwAAAEmCAYAAAAJAaljAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz\nAAALEgAACxIB0t1+/AAAIABJREFUeJzt3Xm8VXW9//HXGw4IBgKBqRwwUhzyeE3F4WZapuaI4E1N\ncyQ1zZ9T3Ua1q2b6y263upXevP7S65iaZgHmVJqV3mQQhwTMUFHAEQQUB5Dj5/fH+h7dbM/Zex3Y\n5+zF2e+nj/VwDd/9XZ+92Hz4rrW+67sUEZiZWXW96h2AmdnawgnTzCwnJ0wzs5ycMM3McnLCNDPL\nyQnTzCwnJ8weQtJ5kq5N8xtLWiapd433MVfSXrWsM8c+T5b0Yvo+Q9egnmWSNqllbPUiaaak3esd\nRyNywswpJYuXJH2gZN0Jku6tY1jtiohnI2JARLTWO5Y1IakP8CNg7/R9Fq1uXenzT9UuutqTdKWk\nC6qVi4iWiLi3G0KyMk6YndMbOGNNK1HGx766DYB+wMx6B1IEkprqHUOj81/azvkB8DVJg9vbKGkX\nSdMkLU3/36Vk272SLpR0P/AGsElad4Gk/02njJMlDZV0naRXUx2jSur4iaR5aduDknbrII5RkkJS\nk6SPp7rbprckzU3lekn6lqQnJS2S9CtJHyyp52hJz6RtZ1c6MJL6S/phKr9U0n2S+qdt49Jp5JL0\nnT9a8rm5kr4m6dH0uRsl9ZO0OfD3VGyJpHtKv1fZcT0hzY+W9KdUz0JJN5aUC0mj0/wgSVdLejnF\n++22f8AkTUix/4ekxZKelrRfhe89V9LXU/yvS7pc0gaSbpf0mqQ/SBpSUv4mSS+kGP8sqSWtPxE4\nEvhG22+hpP5vSnoUeD39mb57aUTSbZJ+WFL/DZKuqPRnZWsgIjzlmIC5wF7ALcAFad0JwL1p/oPA\nYuBooAn4fFoemrbfCzwLtKTtfdK6OcCmwCBgFvBE2k8TcDXwPyUxHAUMTdu+CrwA9EvbzgOuTfOj\ngACayr5DH+BPwPfS8hnAA8AIYB3gv4Hr07atgGXAJ9O2HwErgb06OD6XpO/TTNYS3yV9bnPgdeAz\naf/fSN+5b8lxnQoMT8dwNvCl9r5He98r7fOENH89cDZZQ6AfsGtJuQBGp/mrgYnAwFTnE8DxadsE\n4G3gi+l7nAw8B6jC7+IBstZwM/ASMAPYLsVwD3BuSfnj0n7XAf4TeLhk25Wk31ZZ/Q8DI4H+pb/F\nNL9h2uceZAn3KWBgvf++9NSp7gGsLRPvJcytgaXA+qyaMI8GppZ95q/AhDR/L3B+2fZ7gbNLln8I\n3F6yfGDpX6h2YloMfCzNn0f1hPlz4FagV1qeDexZsn2jlCyagHOAG0q2fQBYQTsJMyWoN9tiKdv2\nb8CvysouAHYvOa5HlWz/d+DS9r5He9+LVRPm1cBlwIh24ghgNFkSXAFsVbLtpJI/xwnAnJJt66bP\nbljhd3FkyfKvgZ+XLJ8G/LaDzw5OdQ9Ky1fSfsI8rr3fYsnywcA8YCEl/0h4qv3kU/JOiojHyJLO\nt8o2DQeeKVv3DFmro828dqp8sWT+zXaWB7QtpFPX2el0bglZq3RYnrglnQTsDhwREe+k1R8GfpNO\nlZeQJdBWstbS8NJ4I+J1oKObLsPIWlNPtrNtleOS9j2PVY/LCyXzb1DynTvpG4CAqekSwHEdxNqH\nVf+syv+c3o0nIt5Is5ViyvVnKKm3pIvSJZBXyRJfW0yVtPe7KTWZ7B+Cv0fEfVXK2hpwwlw955Kd\nspX+JXuOLAGV2pisNdVmtYeGStcrvwF8DhgSEYPJWrrK+dnvAuMj4tWSTfOA/SJicMnULyIWAM+T\nnQa21bEu2eWA9iwE3iK7tFBuleMiSaneBe2Ureb19P91S9Zt2DYTES9ExBcjYjhZq/G/2q5blsX6\nNqv+WZX/OXWVI4DxZGcqg8hazPDen2FHv49qv5sLyf6x20jS59cwRqvACXM1RMQc4Ebg9JLVtwGb\nSzoiXZg/jOw64K012u1AsmuILwNNks4B1qv2IUkjgV8Bx0TEE2WbLwUulPThVHZ9SePTtpuBsZJ2\nldQXOJ8Ofi+p1XgF8CNJw1NL6uOS1kn7PkDSnsq6CX0VWA78b6e+fbafl8kS21FpH8dRkqQlHSpp\nRFpcTJZo3imrozXFdKGkgem7/ytwbWfjWQ0Dyb77IrKk/3/Ltr8IdKqvqKRPAl8AjgGOBX4mqbny\np2x1OWGuvvPJrusBEFkfwbFkCWERWWtwbEQsrNH+7gTuILtB8QxZi67aqRrAnmSn2DfrvTvlbd10\nfgJMAu6S9BrZzYud0/eZCZwC/JKstbkYmF9hP18D/gZMA14Bvk92rfTvZDerfkbWujsQODAiVuT8\n3uW+CHyd7Bi3sGri3RGYImlZ+l5nRPt9L08ja60+BdyXvmN33Fm+muzPbgHZDb4HyrZfDmyVLpH8\ntlplktZLdZ4aEQsi4i+pjv9JLXmrMaWLxmZmVoVbmGZmOTlhmlmPla51PyTpffcSJK2THpSYI2mK\nSh4S6YgTppn1ZGeQ9SBoz/HA4ogYDfyY7Lp7RU6YZtYjpR4TBwC/6KDIeOCqNH8zsGe1m2U94mF+\nNfUP9R1Y7zAKb7uPblzvEKwHmTHjwYURsX6t6uu93ocjVr6Zq2y8+fJMsp4ibS6LiMvKiv0nWW+V\njpJDM6mnSUSslLSUrK9xhz1bekbC7DuQdbb4XL3DKLz7p1xc7xCsB+nfR+VPtq2RWPlm7r/Hbz18\nyVsRsUNH2yWNBV6KiAdVw7FDfUpuZgUhUK98U3WfAMYpG5nrBmAPpQG2SywgPc2WRsAaRMeP/wJO\nmGZWFAJ69c43VRERZ0bEiIgYBRwO3BMRR5UVm0T2dBTAIalMxY7pPeKU3Mx6iC5+QEnS+cD0iJhE\n9lTUNZLmkD2ddni1zzthmllBKO/pdqdE9jqPe9P8OSXr3wIO7UxdTphmVhwFfwTeCdPMikF0SQuz\nlpwwzawg5BammVlubmGameWhXF2G6skJ08yKQfiU3MwsN5+Sm5nl0TX9MGvJCdPMiqOXT8nNzKpz\nP0wzs7x8l9zMLD/fJTczy8mn5GZmOciPRpqZ5ecWpplZTm5hmpnl4Y7rZmb5tL3Tp8CcMM2sIIrf\nwix2dGbWWNrulFebqlajfpKmSnpE0kxJ32mnzARJL0t6OE0nVKvXLUwzK47atTCXA3tExDJJfYD7\nJN0eEQ+UlbsxIk7NW6kTppkVR43ukqf3iy9Li33SVPGd43n4lNzMikHpGmaeCYZJml4ynfj+6tRb\n0sPAS8DvI2JKO3s9WNKjkm6WNLJaiG5hmllhqFfuNtzCiNihUoGIaAW2lTQY+I2krSPisZIik4Hr\nI2K5pJOAq4A9KtXpFuYauPTcI3nm7u8x/aazOizzw28cwmMTz2XqjWey7ZYjujG6YrnrzjvYpmUL\nWrYczQ/+/aL3bV++fDlHHXEYLVuOZrddduaZuXO7P8gCaOTjlL2hQrmmzoiIJcAfgX3L1i+KiOVp\n8RfAmGp1OWGugWsmP8D4Uy7pcPs+u27Fphuvz9bjv8OpF1zPT886vBujK47W1la+fPopTJx8Ow89\nOoubbrie2bNmrVLmyisuZ8jgIcx8fA6nnfEVzj7rm3WKtn4a/jipE1O1qqT1U8sSSf2BzwCPl5XZ\nqGRxHDC7Wr1OmGvg/hlP8srSNzrcPvZT2/DLW6cCMPVvcxk0sD8bDluvu8IrjGlTp7LppqP5yCab\n0LdvXw497HBunTxxlTK3Tp7IkUcfC8BnDz6Ee++5m+y6fePwccrXuszZwtwI+KOkR4FpZNcwb5V0\nvqRxqczpqcvRI8DpwIRqlfoaZhca/qHBzH9h8bvLC15cwvAPDeaFha/WMaru99xzCxgx4r3r6c3N\nI5g6dcr7y4zMyjQ1NbHeoEEsWrSIYcOGdWus9eTjRKdPtzsSEY8C27Wz/pyS+TOBMztTb7e2MCVd\nKemQdtYPl3Rzd8ZiZsXTFdcwa6kQp+QR8VxEvC+Rru2ee2kJIzYc8u5y8waDee6lJXWMqD6GD29m\n/vx57y4vWDCf5ubm95eZl5VZuXIlry5dytChQ7s1znrzcWrwhCnpmNTH6RFJ16TVn5T0v5Keamtt\nShol6bE0P0HSREn3SvqHpHO7Msau9Ls//Y0jxu4EwE7/NIpXl73ZcKfjADvsuCNz5vyDuU8/zYoV\nK7jpxhs4YOy4VcocMHYc111zFQC3/PpmPvXpPer6F6MeGv04SUK98k310mXXMCW1AN8GdomIhZI+\nCPyI7GLsrsCWwCSgvVPxnYCtgTeAaZJ+FxHTy+o/Ecg6q/YZ0FVfo6KrvjeB3cZsxrDBA5hzx3f5\n7qW30acpG23lFzffxx33zWSfXVuYOelc3njrbU4679q6xFlvTU1N/PgnF3PgAfvQ2trKsROOY6uW\nFs4/7xy2H7MDYw8cx4Tjjue4CUfTsuVohgz5INdcd0O9w+52Pk61u4bZVdRVd9gknQZsGBFnl6y7\nkuxu1XVp+bWIGChpFHBrRGwtaQLZM6DHpDLnA69ExH92tK9e634o1tnic13yPXqSxdMurncI1oP0\n76MHq3Ue74ymoZvEevtfkKvs4muPrOm+86rHXfLlJfMd/XNSnsV7Sr8JM6ug6C3MrryGeQ9wqKSh\nAOmUPK/PSPpg6nB6EHB/VwRoZgVSw47rXaXLWpgRMVPShcCfJLUCD3Xi41OBXwMjgGvLr1+aWc9U\n9BZml56SR8RVZA+0d7R9QPr/XLKbPG3mR8RBXRmbmRWLEL3yD75RF37Sx8yKo9gNzOIlzIi4Eriy\nzmGYWXdTg5+Sm5l1hhOmmVlOTphmZjmI+j4nnocTppkVg6jrc+J5OGGaWWG4hWlmlpMTpplZXsXO\nl06YZlYcRW9hFvs5JDNrGHlHW8+TVCX1kzQ1DV4+U9J32imzjqQbJc2RNCUNM1mRE6aZFUYNX1Gx\nnGxc3Y8B2wL7SvrnsjLHA4sjYjTwY+D71Sp1wjSzwqjVKyoisywt9klT+bi643lvcKCbgT1VJRs7\nYZpZYXSihTlM0vSS6cR26uot6WHgJbI3PUwpK9IMzAOIiJXAUqDiG+V808fMiqFzg28srPaKioho\nBbaVNBj4jaStI+KxNQnRLUwzKwQBUr6pMyJiCfBHYN+yTQuAkQCSmoBBwKJKdTlhmllB1PQu+fqp\nZUl61c1ngMfLik0Cjk3zhwD3RJW3QvqU3MwKo4bdMDcCrpLUm6xh+KuIuDW9hXZ6REwCLgeukTQH\neAU4vFqlTphmVgyCXjUafCMiHgW2a2f9OSXzbwGHdqZeJ0wzKwRRu4TZVZwwzawwCv5kpBOmmRVH\n0Z8ld8I0s2JYjS5D3c0J08wKIeuHWeyM6YRpZgXhd/qYmeXmu+RmZnn4GqaZWT6+hmlm1gkFz5dO\nmGZWHG5hmpnlVPB82TMS5nYf3Zj7p1xc7zAKb8iOp9Y7hLXC4mn+LdWDajj4RlfpEQnTzHoC98M0\nM8ut4PnSCdPMisMtTDOzPNxx3cwsn7Wh47pfgmZmhdGrl3JN1UgaKemPkmZJminpjHbK7C5pqaSH\n03ROe3WVcgvTzAqjhi3MlcBXI2KGpIHAg5J+HxGzysr9JSLG5q3UCdPMiqGG1zAj4nng+TT/mqTZ\nQDNQnjA7xafkZlYI6tx7yYdJml4yndhhvdIosjdITmln88clPSLpdkkt1WJ0C9PMCqMTLcyFEbFD\n9fo0APg18OWIeLVs8wzgwxGxTNL+wG+BzSrV5xammRVGLynXlIekPmTJ8rqIuKV8e0S8GhHL0vxt\nQB9JwyrG1/mvZGbWNaR8U/V6JOByYHZE/KiDMhumckjaiSwfLqpUr0/JzawQJOhdu8E3PgEcDfxN\n0sNp3VnAxgARcSlwCHCypJXAm8DhERGVKu0wYUpar9IH27keYGa2RmrVrSgi7iPrC1+pzMVAp4am\nqtTCnAlE2U7bloOUqc3MaqXgD/p0nDAjYmR3BmJmjU1kXYuKLNdNH0mHSzorzY+QNKZrwzKzRtRL\n+aa6xVetgKSLgU+TXUAFeAO4tCuDMrMGlLPTej0H6Mhzl3yXiNhe0kMAEfGKpL5dHJeZNRhR07vk\nXSJPwnxbUi+yGz1IGgq806VRmVlDKvpNnzzXMC8h6y2/vqTvAPcB3+/SqMysIa31p+QRcbWkB4G9\n0qpDI+Kxrg3LzBpN3qd46invkz69gbfJTsv9OKWZdYm8z4nXS5675GcD1wPDgRHALyWd2dWBmVnj\nUc6pXvK0MI8BtouINwAkXQg8BHyvKwMzs8ZT9Hf65EmYz5eVa0rrzMxqRtLa261I0o/Jrlm+AsyU\ndGda3huY1j3hmVkjKXgDs2ILs+1O+EzgdyXrH+i6cMyska21p+QRcXl3BmJmjU3U9znxPPLcJd9U\n0g2SHpX0RNvUHcGtDe668w62admCli1H84N/v+h925cvX85RRxxGy5aj2W2XnXlm7tzuD7LOLj33\nSJ65+3tMv+msDsv88BuH8NjEc5l645lsu+WIboyuWBr991T0jut5+lReCfwP2T8A+wG/Am7swpjW\nGq2trXz59FOYOPl2Hnp0FjfdcD2zZ636Fs8rr7icIYOHMPPxOZx2xlc4+6xv1ina+rlm8gOMP+WS\nDrfvs+tWbLrx+mw9/jucesH1/PSsw7sxuuLw76n43YryJMx1I+JOgIh4MiK+TZY4G960qVPZdNPR\nfGSTTejbty+HHnY4t06euEqZWydP5MijjwXgswcfwr333E2VUfB7nPtnPMkrS9/ocPvYT23DL2+d\nCsDUv81l0MD+bDis4oD/PVKj/57aXlGRZ6qXPAlzeRp840lJX5J0IDCwi+NaKzz33AJGjHhvnOXm\n5hEsWLDg/WVGZmWamppYb9AgFi2q+J6lhjP8Q4OZ/8Lid5cXvLiE4R8aXMeI6sO/p55xSv4V4APA\n6WQvFvoicFxndiJplCQ/f25mFdXwrZEjJf1R0ixJMyWd0U4ZSfqppDnpHs321erNM/jGlDT7Gu8N\nImzA8OHNzJ8/793lBQvm09zc/P4y8+YxYsQIVq5cyatLlzJ06NDuDrXQnntpCSM2HPLucvMGg3nu\npSV1jKg+Gv33JPK/czyHlcBXI2KGpIHAg5J+HxGlF4X3AzZL087Az9P/O9RhC1PSbyTd0tG0ut9C\n0iaSHpK0s6QfSJqWsvtJafvVkg4qKX+dpPGru7+utMOOOzJnzj+Y+/TTrFixgptuvIEDxo5bpcwB\nY8dx3TVXAXDLr2/mU5/eo/B9zbrb7/70N44YuxMAO/3TKF5d9iYvLGy8l5I2/O8pZ+syz9eNiOcj\nYkaafw2YDTSXFRsPXB2ZB4DBkjaqVG+lFmanXj+Zh6QtgBuACWSZfGlE7ChpHeB+SXeRvXz9K8Bv\nJQ0CdgGObaeuE4ETAUZuXJ8XWDY1NfHjn1zMgQfsQ2trK8dOOI6tWlo4/7xz2H7MDow9cBwTjjue\n4yYcTcuWoxky5INcc90NdYm1nq763gR2G7MZwwYPYM4d3+W7l95Gn6beAPzi5vu4476Z7LNrCzMn\nncsbb73NSeddW+eI68O/p051XB8maXrJ8mURcVkHdY4CtgOmlG1qBuaVLM9P6zp89FvdcYctBTwF\nWAx8NiJmSboZ2IbsHUEAg4CTIuIuSTOB3YGDgdER8bVK9Y8Zs0PcP2V6pSIGDNnx1HqHsFZYPK3m\nbYUeqX8fPRgRO9Sqvg+N3joO+8FNucpe/Nmtcu1b0gDgT8CFEXFL2bZbgYvSO8yRdDfwzYjoMJnk\nHQ+zFpYCzwK7ArPIulOd1tZlqczVwFHA4cAXui1CM6ubWr/TR1IfsrdFXFeeLJMFQOnrxEekdR3q\nzsGAVwD/Ahwj6QjgTuDk9KWQtLmkD6SyVwJfBii7SGtmPVitXrOr7Nz+cmB2RPyog2KTyPKRJP0z\n2SXCiiOx5W5hSlonIpbnLd+eiHhd0ljg98B3yVqaM9KXexk4KJV7UdJs4Ldrsj8zW3tkN3Rq1sL8\nBFmvnr9JejitOwvYGCAiLgVuA/YH5pBdGqx6Nls1YUraiSxTDwI2lvQx4ISIOC1v5BExF9g6zS8B\ndkybJqUvUb7Pdclu9V+fdx9mtvar1Rl5ui5ZsbbIbuCc0pl685yS/xQYCyxKO3kE+HRndtIZkvYi\n6wLws4hY2lX7MbPiqVW3oq6S55S8V0Q8U9ZUbu2ieIiIPwAf7qr6zayYsuHdit2nNE/CnJdOy0NS\nb+A0wMO7mVnN9S52vsyVME8mOy3fGHgR+ENaZ2ZWM1JNH43sEnmeJX+JrD+kmVmXKni+zHWX/P+R\nvfxsFRFxYpdEZGYNq+ivqMhzSv6Hkvl+ZJ3P53VQ1sxstfSImz4RscrrKCRdA9zXZRGZWcMqeL5c\nrWfJPwJsUOtAzKzBCXoXPGPmuYa5mPeuYfYCXgG+1ZVBmVnjWRtes1sxYaZnvD/GeyN4vBM95Y1L\nZlY4RU+YFR+NTMnxtohoTZOTpZl1mZ7wErSHJW3X5ZGYWUNrOyWvxfBuXaXDU3JJTRGxkmxo92mS\nngReJ/teERFV37BmZpZbnQfWyKPSNcypwPbAuAplzMxqZm3uhymAiHiym2IxswaWvaKi3lFUVilh\nri/pXzvaWGHYdzOz1SB6VR7zt+4qJczewACqjFpsZlYLYu2+hvl8RJzfbZGYWWOr8x3wPCpdMSh4\n6GbW0/RKY2JWm6qRdIWklyQ91sH23SUtlfRwms7JE1+lFuaeeSowM6uFGp+SXwlcDFxdocxfImJs\nZyrtMGFGxCudqcjMbE31rtE5eUT8WdKomlRWouA38c2sUYgsIeWZgGGSppdMqzOg+cclPSLpdkkt\neT6wOsO7mZnVnujMc+ILI2KHNdjbDODDEbFM0v7Ab4HNqn3ILUwzKwzlnNZURLwaEcvS/G1AH0nD\nqn3OLUwzK4TufEWFpA2BFyMi0mvEewGLqn3OCdPMCqNW6VLS9cDuZNc65wPnAn0AIuJS4BDgZEkr\ngTeBw/MMX+mEaWaFUasGZkR8vsr2i8m6HXWKE6aZFYLQ2v9OHzOz7lLP0dTzcMI0s8Iodrp0wmwo\ni6d1+pJNQ/rqpFn1DqExda4fZl04YZpZIbQ96VNkTphmVhhuYZqZ5VT08TCdMM2sELJT8mJnTCdM\nMyuMgp+RO2GaWVEIuYVpZpaPW5hmZjn4GqaZWV6CXgXviOmEaWaF4WuYZmY5ZAMI1zuKypwwzaww\n3MI0M8vJd8nNzHIqeguz4PekzKxRtF3DzDNVrUu6QtJLkh7rYLsk/VTSHEmPSto+T4xOmGZWDBK9\nck45XAnsW2H7fmTvId8MOBH4eZ5KnTDNrDBq9V7yiPgz8EqFIuOBqyPzADBY0kbV6vU1TDMrhO58\nLznQDMwrWZ6f1j1f6UNOmGZWGJ1Il8MkTS9ZviwiLqt5QGWcMM2sOPJnzIURscMa7GkBMLJkeURa\nV5GvYZpZYSjnfzUwCTgm3S3/Z2BpRFQ8HQe3MM2sQGr1aKSk64HdyU7d5wPnAn0AIuJS4DZgf2AO\n8AbwhTz1OmGaWXHUKGFGxOerbA/glM7W64RpZoWQdRkq9pM+TphmVgzys+RmZrkVPF86YZpZgRQ8\nYzphmllB+K2RZma5rA0jrrvj+hq668472KZlC1q2HM0P/v2i921fvnw5Rx1xGC1bjma3XXbmmblz\nuz/IAvBxqu6o7Tfiov035+w9N+mwzKHbbMB5e4/mrD03YeTgft0YXTep1egbXcQJcw20trby5dNP\nYeLk23no0VncdMP1zJ41a5UyV15xOUMGD2Hm43M47YyvcPZZ36xTtPXj45TPA88s5ZL/fbbD7S0b\nDGD9Aetw3l1z+OWM5zl826qD66x1uvFJn9XihLkGpk2dyqabjuYjm2xC3759OfSww7l18sRVytw6\neSJHHn0sAJ89+BDuvedusj6zjcPHKZ85i97g9RWtHW7fZvhApjy7BIC5i9+kf59erNevZ11Vk/JN\n9eKEuQaee24BI0a89/x+c/MIFixY8P4yI7MyTU1NrDdoEIsWLerWOOvNx6k2BvVrYsmbb7+7vOTN\nlQzuaQkz51QvPetom9naq97ZMAe3MNfA8OHNzJ//3hikCxbMp7m5+f1l5mVlVq5cyatLlzJ06NBu\njbPefJxqY+lbKxncv8+7y4P7N7HkrZV1jKi22gYQrtErKrpElydMSaMkPS7pOkmzJd0saV1JYyT9\nSdKDku5sGx5e0r2Svi9pqqQnJO3W1TGurh123JE5c/7B3KefZsWKFdx04w0cMHbcKmUOGDuO6665\nCoBbfn0zn/r0Hqjoz3/VmI9TbTz6/GvsvPFgAEYN6c+bb7/Dqz0oYYJPydtsARwfEfdLuoJslJB/\nAcZHxMuSDgMuBI5riysidpK0P9mwTHuVVyjpRLKXFzFy44274zu8T1NTEz/+ycUceMA+tLa2cuyE\n49iqpYXzzzuH7cfswNgDxzHhuOM5bsLRtGw5miFDPsg1191Ql1jryccpny/s2Mxm66/LgL5NXLDf\nZvxu1sv0Th0T73t6MTNfWEbLBgM4b+/RrGh9h2sffK7OEXeBgv8bqa6+EylpFPDniNg4Le8BnAXs\nBDyVivUGno+IvSXdC5ydkusGwP0RMbrSPsaM2SHunzK9UhGz3L46aVb1QsZ/Hdzy4BqOer6KrT+2\nfdx8x325yn50+Adquu+8uquFWZ6VXwNmRsTHOyi/PP2/Fd+YMmsYRb8K0103fTaW1JYcjwAeANZv\nWyepj6SWborFzAqq6Ncwuyth/h04RdJsYAjwM+AQ4PuSHgEeBnbppljMrKgKnjG763R3ZUQcVbbu\nYeCT5QUjYveS+YXAqC6NzMwKQerW95KvFvfDNLPCqGUDU9K+kv4uaY6kb7WzfYKklyU9nKYTqtXZ\n5S3MiJgLbN3V+zGzHqB2b43sDVwCfAaYD0yTNCkiyrtA3BgRp+at1y1MMyuIvGMV5cqqOwFzIuKp\niFgB3ACMX9MInTDNrDA6MVrRMEnTS6YTy6pqBuaVLM9P68odLOnR9ATiyHa2r8J9HM2sEDp5A3xh\nDTquTwZ33Kw3AAAKn0lEQVSuj4jlkk4CrgL2qPQBtzDNrDAk5ZpyWACUthhHpHXviohFEdH2kMwv\ngDHVKnXCNLPCqOEAwtOAzSR9RFJf4HBg0qr7UumQ9eOA2dUq9Sm5mRVGrXphRsRKSacCd5KNVXFF\nRMyUdD4wPSImAadLGgesBF4BJlSr1wnTzIqhxq+fiIjbgNvK1p1TMn8mcGZn6nTCNLMCKfaTPk6Y\nZlYIovijFTlhmllh9HLCNDPLp57vHM/DCdPMiqPY+dIJ08yKo+D50gnTzIqhE53S68YJ08wKw9cw\nzczyKna+dMI0s+JwtyIzs1xyDw5cN06YZlYIa8OTPh7ezcwsJ7cwzawwit7CdMI0s8LwNUwzsxwk\n3yU3M8vPCdPMLB+fkpuZ5VT0mz7uVmRmhaGcU666pH0l/V3SHEnfamf7OpJuTNunSBpVrU4nTDMr\njhplTEm9gUuA/YCtgM9L2qqs2PHA4ogYDfwY+H61ep0wzawwlPO/HHYC5kTEUxGxArgBGF9WZjxw\nVZq/GdhTqnxRoEdcw5wx48GF/fvomXrHUWYYsLDeQawFfJzyKeJx+nAtK3toxoN3rttXw3IW7ydp\nesnyZRFxWclyMzCvZHk+sHNZHe+WSe8xXwoMpcJx7hEJMyLWr3cM5SRNj4gd6h1H0fk45dMIxyki\n9q13DNX4lNzMeqIFwMiS5RFpXbtlJDUBg4BFlSp1wjSznmgasJmkj0jqCxwOTCorMwk4Ns0fAtwT\nEVGp0h5xSl5Ql1UvYvg45eXj1AnpmuSpwJ1Ab+CKiJgp6XxgekRMAi4HrpE0B3iFLKlWpCoJ1czM\nEp+Sm5nl5IRpZpaTE6aZWU5OmFYYkj5Q7xjMKnHCtEKQ9FHghDTv32VO1R7ls9ryD7MLtf2YJW0i\n6aNpQIBVttm7hgFflrRFRLxT72CKStKeko6XdAhARIR/S93HCbMLpR/z/sBE4BrgPyX9S9u2ugZX\nEJL6SeoVEX8Bfg6Mk9TLrcz3k7QzWd/BUcDXJF0ETprdyT/KLiTpn4CvAJ8FPg08AXxc0tZ1Dawg\n0nH4PvBtSQOAh8iG4iIi3nESeI+kHYFDgdMj4t+Aw4C9JH0P/A9wd3HCrCFJG0q6OLWQBgBHAVsD\nvSPiNbIhpj4M7FPPOOup5DKFgCeBu4D1gFuADYADgdPASaDMzsA4YLSkdSLiGbJ/iA+S9MP6htY4\n/Ghkbb0C/Izsgf4XgIuBDwGnSvpRRDwl6Q5gW0lNEbGyjrHWRcllin2Bl4AbI+J3ksYC/YDngW3q\nGWORSNoEeCEiLpb0PHASMEXS1Ih4VtK+ZANLWDfwo5FdQNIvgM2Bz5D9mL8E7ApcC3wROCsibqtf\nhPUjaTeya5WHAr8B/gx8qe1Gj6TNgZuAr0fEXXULtAAk7Qd8F7gd2J5swNtTyH5XPwTui4i36xdh\n4/EpeQ2UnWZClhRnk51mzgd+Aswha1V9NyJuK71j3tO1fVdJ/cn+4fgW8EFgKdnxeCdtIyKeIDtN\n37xO4RZCep3ChcDnyI7TMKBfRPwEuBv4NjCgfhE2JifMNSBpkKQN0mnmfsAFkr4dmZPITstvIhvB\n+TzgQbJh8EdGRGv9Iu8e6TouEdEq6RPAf5D9A3IS2TtUPhcR8yQdCXw5fWYYsCFwT32iLozlwC+A\nLciS5hERsUzSLhHxY+ALEbG4rhE2ICfM1ZSeSrkI+KykvYEfAU8B4yXdAhARxwOvARMj4kngD8Bz\nwIr6RN19JK0L3CapbbzBJcBiYDrZQK2XAa9J2g74JvAIQEQsBE6IiFndH3X9SfqEpGOAHYDvkP2u\n9oiIJyV9Ejhb0kYR8WxdA21Qvoa5BiTtQ3YtbhnweERcmtZPBZ6JiEPT8sci4pE03y8i3qpXzN1J\n0kHAOWRdh2YCJ0bE6ZIOBPYkaz31Bi6JiImpP2bDdlqXtAtZP8sHyW5+7QH0JWuZvwmcCZwXERPr\nFmSDc8JcDZJ6p9PMjwFnkHWLWQxckLp7IGkm8GREjEvLDZkMJB1Adi3uEWAd4H/ITjc3IXtFwMyI\nmC9JjdyNSNJOZP+wnBkRD0jaFBgLfJys98Ac4O6IuL3Rj1U9uVvRakjJcjxwLtkd8AHAF8iuT94V\nEfMjoiVdt2v7TMMlS4DUZagV+CnZG/n+DOxGdszOj4j5qVyjJ4BBwCfJWpUPAM8Cc4GREfG1tkJO\nlvXlhLka0s2MCcCpETE1rVuX7L0gfSXdERFzI+L+OoZZGBFxh6QvkbWgnoiI/6p3TEUTEb+X9Fng\nh5KejojrJS0BPiVpA+CldDPRybKOnDBXTwDrk7p1pNPtW1MfwjHA5HoGV0QRcY+kC4Gfpmu/zzdi\nx/1K0nXcd4DrJB0MvEPW7erFOodmie+Sr4aIeB24EdhF0kdTP8KPk93ZvDQiyl/naUBE/Bb4VETM\nc7JsX0RMJnukdjQwLSImKalzaIZv+qw2Sc1k/Qk/DdxH9sa5Uxr1CR6rrdRV7QqywTZuqXc8lnHC\nXAOpL+aOZINGzI2IKXUOyXoQSZ8h62nxVL1jsYwTpplZTr6GaWaWkxOmmVlOTphmZjk5YZqZ5eSE\naWaWkxNmg5HUKulhSY9Juik90rm6de0u6dY0P07StyqUHSzp/6zGPs6T9LW868vKXKn0Otqc+xol\n6bHOxmiNwwmz8bwZEdtGxNZk43J+qXRjeqik07+LiJgUERdVKDIY6HTCNCsSJ8zG9heytxCOkvR3\nSVcDjwEjJe0t6a+SZqSWaNtz8/tKelzSDLK3FpLWT5B0cZrfQNJvJD2Spl3IBlveNLVuf5DKfV3S\nNEmPSvpOSV1nS3pC0n1kY2ZWJOmLqZ5HJP26rNW8l6Tpqb6xqXxvST8o2fdJa3ogrTE4YTYoSU3A\nfsDf0qrNgP+KiBbgdbJ3xuwVEduTjZL+r5L6Af+P7FW4Y8heJdGenwJ/ioiPkb28aybZe3yeTK3b\nr6dH/zYDdgK2BcZI+qSkMWSPmW4L7E/2JFU1t0TEjml/s4HjS7aNSvs4ALg0fYfjgaURsWOq/4uS\nPpJjP9bgPFpR4+kv6eE0/xeyEb6Hk40Q/0Ba/8/AVsD9acyHvsBfgS2BpyPiHwCSrgVObGcfewDH\nQDZ2KLBU0pCyMnun6aG0PIAsgQ4EfhMRb6R9TMrxnbaWdAHZaf8A4M6Sbb9KY5H+Q9JT6TvsDWxT\ncn1zUNr3Ezn2ZQ3MCbPxvBkR25auSEnx9dJVwO8j4vNl5Vb53BoS8L2I+O+yfXx5Neq6EjgoIh6R\nNAHYvWRb+bO/kfZ9WkSUJlYkjVqNfVsD8Sm5tecB4BOSRkM2yEga6/NxYFR6fQLA5zv4/N3Ayemz\nvSUNInsZ3MCSMncCx5VcG22W9CGyEdkPktRf0kCy0/9qBgLPS+oDHFm27VBJvVLMmwB/T/s+OZVH\n0uZpIBWzitzCtPeJiJdTS+16Seuk1d+OiCcknQj8TtIbZKf0A9up4gzgMknHA63AyRHxV0n3p247\nt6frmB8F/ppauMuAoyJihqQbyd4B9BIwLUfI/wZMAV5O/y+N6VlgKtl7l74UEW9J+gXZtc0ZaZzJ\nl4GD8h0da2QercjMLCefkpuZ5eSEaWaWkxOmmVlOTphmZjk5YZqZ5eSEaWaWkxOmmVlO/x9wIoWX\nlcekzQAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fe025079150>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "import cv2\n",
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "from os import listdir, path\n",
    "from sklearn.metrics import confusion_matrix\n",
    "from sklearn.metrics import accuracy_score\n",
    "import itertools\n",
    "%matplotlib inline\n",
    "\n",
    "imagesnames = sorted(listdir('training_set')) #Get list with the images directions for the training set\n",
    "del imagesnames[0] #Delete the Binarized Images folder direction\n",
    "vect = [] #Array for the images features\n",
    "\n",
    "for name in imagesnames:\n",
    "    image = cv2.imread(\"training_set/\"+name,0) #Read image\n",
    "    image = cv2.medianBlur(image,7) #Apply filter\n",
    "    th1 = cv2.adaptiveThreshold(image,255,cv2.ADAPTIVE_THRESH_MEAN_C,cv2.THRESH_BINARY_INV,255,2) #Use the adaptive threshold technique to get a binary image from the original\n",
    "    kernel = np.ones((5,5),np.uint8) #Define structuring element\n",
    "    img=cv2.dilate(th1,kernel,iterations = 1) #Dilate the binary image\n",
    "    HM=cv2.HuMoments(cv2.moments(img)).flatten() #Get image features through HuMoments\n",
    "    vect.append(HM) #Add the features to the array\n",
    "\n",
    "# Feature set containing values of training data\n",
    "trainData = np.asarray(vect).astype(np.float32)\n",
    "\n",
    "# Label each image from the training set\n",
    "responses = [0,0,0,0,0,0,0,0,0,1,1,1,1,1,1,1,2,2,2,2,2] #responses must be integers\n",
    "dictionary = {0: 'chip', 1: 'key', 2: 'pen'} #the meaning of each integer\n",
    "responses = np.asarray(responses).astype(np.float32)\n",
    "\n",
    "#Build predictable model \n",
    "knn = cv2.KNearest() #Classifier\n",
    "knn.train(trainData,responses) #Train the model\n",
    "\n",
    "tsImages = sorted(listdir('testing_set')) #Get list with the images directions for the testing set\n",
    "del tsImages[0] #Delete the Binarized Images folder direction\n",
    "\n",
    "'''Plot Confusion Matrix taken from: http://scikit-learn.org/stable/auto_examples/model_selection/plot_confusion_matrix.html'''\n",
    "def plot_confusion_matrix(cm, classes,\n",
    "                          normalize=False,\n",
    "                          title='Confusion matrix',\n",
    "                          cmap=plt.cm.Blues):\n",
    "    \"\"\"\n",
    "    This function prints and plots the confusion matrix.\n",
    "    Normalization can be applied by setting `normalize=True`.\n",
    "    \"\"\"\n",
    "    plt.imshow(cm, interpolation='nearest', cmap=cmap)\n",
    "    plt.title(title)\n",
    "    plt.colorbar()\n",
    "    tick_marks = np.arange(len(classes))\n",
    "    plt.xticks(tick_marks, classes, rotation=45)\n",
    "    plt.yticks(tick_marks, classes)\n",
    "\n",
    "    if normalize:\n",
    "        cm = cm.astype('float') / cm.sum(axis=1)[:, np.newaxis]\n",
    "        \n",
    "    else:\n",
    "        cm = cm\n",
    "\n",
    "    thresh = cm.max() / 2.\n",
    "    for i, j in itertools.product(range(cm.shape[0]), range(cm.shape[1])):\n",
    "        plt.text(j, i, cm[i, j],\n",
    "                 horizontalalignment=\"center\",\n",
    "                 color=\"white\" if cm[i, j] > thresh else \"black\")\n",
    "\n",
    "    plt.tight_layout()\n",
    "    plt.ylabel('True label')\n",
    "    plt.xlabel('Predicted label')\n",
    "    \n",
    "def KNNprediction(image, kVector):\n",
    "    plt.figure();plt.subplot(1,2,1);plt.imshow(image,cmap='gray');plt.axis('off');plt.title('Original Image')\n",
    "    image=np.uint8(image)\n",
    "    image = cv2.medianBlur(image,7) #Apply filter\n",
    "    th1 = cv2.adaptiveThreshold(image,255,cv2.ADAPTIVE_THRESH_MEAN_C,cv2.THRESH_BINARY_INV,255,2) #Binary image for feature extraction\n",
    "    kernel = np.ones((5,5),np.uint8) #Define structuring element\n",
    "    img=cv2.dilate(th1,kernel,iterations = 1) #Dilate the binary image\n",
    "    plt.subplot(1,2,2);\n",
    "    plt.imshow(img, cmap=\"gray\"); plt.title(\"Binarized Image\");plt.axis('off')\n",
    "    newcomer=cv2.HuMoments(cv2.moments(img)).flatten() #Extract image features\n",
    "    newcomer = np.asarray([newcomer]).astype(np.float32)\n",
    "    plt.show()\n",
    "    res = []\n",
    "    for k in range(len(kVector)):\n",
    "        ret, results, neighbours ,dist = knn.find_nearest(newcomer, kVector[k]) #Model prediction\n",
    "        a = []\n",
    "        print \"Using \"+str(kVector[k])+\" neighbours: \" #Number of K neighbors\n",
    "        \n",
    "        print \"Result: \", dictionary[results[0][0]].capitalize() #Predicted class\n",
    "        res.append(dictionary[results[0][0]])\n",
    "        for j in range(len(neighbours[0])):\n",
    "            a.append(dictionary[neighbours[0][j]]) \n",
    "        print \"Neighbours: \", a #List of neighbors\n",
    "        print \"Distance: \", dist #Distance to the neighbors\n",
    "        print \"\\n\"\n",
    "    return res\n",
    "    \n",
    "kVector = [1,3,9] #Ks to use with the model\n",
    "predictions = []\n",
    "for i in range(len(tsImages)):\n",
    "    # Newcomer\n",
    "    print \"Performing image recognition on file: \"+tsImages[i]\n",
    "    image = cv2.imread(\"testing_set/\"+tsImages[i],0) #Read newcomer image\n",
    "    res = KNNprediction(image, kVector) #Call the function\n",
    "    predictions.append(res)\n",
    "\n",
    "predictions = np.array(predictions)\n",
    "#Get arrays with the predictions for each K\n",
    "y_pred_1 = predictions[:,0] #K = 1\n",
    "y_pred_2 = predictions[:,1] #K = 3\n",
    "y_pred_3 = predictions[:,2] #K = 9\n",
    "\n",
    "print \"--------------------------------------------\"\n",
    "#Confusion matrix for K=1\n",
    "y_true = ['chip','chip','chip','chip','key','key','key','key','pen','pen'] #Expected labels\n",
    "print \"Acuracy and confusion matrix for K=1\"\n",
    "cnf_matrix_1=confusion_matrix(y_true, y_pred_1) #Confusion matrix\n",
    "# Plot normalized confusion matrix\n",
    "plt.figure()\n",
    "print \"Accuracy: \",accuracy_score(y_true, y_pred_1) #Accuracy of the K-NN classifier\n",
    "plot_confusion_matrix(cnf_matrix_1, classes=['chip','key','pen'], normalize=True,\n",
    "                      title='Normalized confusion matrix')\n",
    "plt.show()\n",
    "\n",
    "#Confusion matrix for K=3\n",
    "print \"Acuracy and confusion matrix for K=3\"\n",
    "cnf_matrix_2=confusion_matrix(y_true, y_pred_2) #Confusion matrix\n",
    "# Plot normalized confusion matrix\n",
    "plt.figure()\n",
    "print \"Accuracy: \",accuracy_score(y_true, y_pred_2) #Accuracy of the K-NN classifier\n",
    "plot_confusion_matrix(cnf_matrix_2, classes=['chip','key','pen'], normalize=True,\n",
    "                      title='Normalized confusion matrix')\n",
    "plt.show()\n",
    "\n",
    "#Confusion matrix for K=9\n",
    "print \"Acuracy and confusion matrix for K=9\"\n",
    "cnf_matrix_3=confusion_matrix(y_true, y_pred_3) #Confusion matrix\n",
    "# Plot normalized confusion matrix\n",
    "plt.figure()\n",
    "print \"Accuracy: \",accuracy_score(y_true, y_pred_3) #Accuracy of the K-NN classifier\n",
    "plot_confusion_matrix(cnf_matrix_3, classes=['chip','key','pen'], normalize=True,\n",
    "                      title='Normalized confusion matrix')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "collapsed": true,
    "deletable": true,
    "editable": true
   },
   "source": [
    "<h3>- Descripción de la función desarrollada:</h3>  \n",
    "<p style=\"text-align:justify\">Para el desarrollo de este problema se desarrolló la función KNNprediction(). Esta función recibe como parámetros de entrada una imagen, la cual corresponde a un objeto al cual se le desea realizar reconocimiento, y además recibe el número de vecinos con los cuales se desea llevar a cabo la predicción, en forma de vector. Esta función realiza internamente la binarización de la imagen de entrada y realiza una gráfica de la misma. A partir de esta imagen se obtienen sus características a través del método de los momentos de Hu, mediante los cuales se construye el modelo de predicción de los K-vecinos más cercanos. Finalmente, se presentan los resultados de la predicción para los valores de K contenidos en el vector ingresado y se retornan como salida de la funcion. Cada uno de estos resultados es guardado en tres vectores distintos para el caso de 1, 3 y 9 vecinos. Estos vectores son utilizados posteriormente para calcular la precisión del clasificador y para construir cada una de las matrices de confusión que permitirán evaluar el desempeño del modelo de reconocimiento de objetos.  </p>\n",
    "<h3>- Descripción de las funciones de librerías utilizadas:</h3>\n",
    "<p style=\"text-align:justify\">Con el fin de binarizar la imagen de entrada y extraer sus características mediante los momentos de Hu, primero se debe suavizar la imagen en escala de grises utilizando la función medianBlur() de la librería OpenCV, la cual tiene como parámetros de entrada la imagen en escala de grises y el tamaño del kernel que se usará. Este kernel recorre todos los pixeles de la imagen mientras toma la mediana de todos los píxeles bajo su área y sustituye el elemento central por este valor mediano, logrando suavizar adecuadamente la imagen. Seguidamente se binariza cada imagen utilizando la técnica de umbral adaptativo con ayuda de la función adaptiveThreshold() de OpenCV, la cual calcula el umbral para pequeñas regiones de la imagen en lugar de la imagen entera, obteniendo diferentes umbrales para diferentes regiones de la misma imagen y da mejores resultados para imágenes con iluminación variable. Esta función recibe seis parámetros de entrada: la imagen, el máximo valor de umbral, el método adaptativo, el tipo de umbral, el tamaño de un vecindario de píxeles que se utiliza para calcular el valor de umbral para el píxel y una constante substraída de la media o de la media ponderada. </p>\n",
    "<p style=\"text-align:justify\">Luego de esto se debe dilatar la imagen binarizada con el fin de conseguir mejores resultados. Para esto se utilizaron distintas funciones, siendo primera la función ones(), con la que se puede obtener una matriz de unos que servirá como elemento estructurante para la dilatación. Esta función recibe el tamaño y tipo de datos deseado para la matriz, que en este caso será de tamaño 3x3 y con datos enteros sin signo de 8 bits (uint8). Luego se utilizó la función dilate(), la cual realiza la operación morfológica de dilatación y recibe como parámetros de entrada la imagen a la cual se la aplicará la operación morfológica, el elemento estructurante y el número de iteraciones. Una vez completado este proceso se tiene la imagen binarizada deseada para la extracción de características.</p>\n",
    "<p style=\"text-align:justify\">Para hallar el conjunto de los siete descriptores invariantes a rotación, escala y traslación que cuantifican la forma de un objeto, es decir los momentos de Hu, se ejecutaron distintas funciones. Una de estas fue moments(), la cual recibe una imagen como parámetro de entrada y calcula todos los momentos hasta el tercer orden de un polígono o forma rasterizada. También se utilizó la función HuMoments(), que recibe como parámetro de entrada los momentos calculados con moments(), para calcular los siete invariantes momentos de Hu.</p>\n",
    "<p style=\"text-align:justify\">El proceso anterior se repite para cada una de las imágenes del conjunto de entrenamiento, mientras se almacenan en un vector de vectores las características de cada imagen extraidas mediante los momentos de Hu. Una vez guardadas las características de todas las imágenes, se debe crear otro vector con las etiquetas correspondientes a cada imagen. Este vector debe ser del mismo tamaño que el vector de vectores de características, claramente. Por requerimientos de las funciones usadas para la clasificación por K-Nearest Neighbors (K-NN), las etiquetas deben ser números enteros. Sin embargo, se creo un diccionario que relaciona los números usados con las etiquetas reales de las imágenes. El siguiente paso es construir el modelo predictor. Para ello se define el clasificador con la función Knearest() de OpenCV, la cual retorna un objeto del tipo clasificador K-NN. Seguidamente se entrena este clasificador utilizando su metodo train(), que recibe dos vectores como entrada: El vector con las características de las imágenes del conjunto de entrenamiento y el vector con las etiquetas de dichas imágenes. </p>\n",
    "<p style=\"text-align:justify\">Una vez entrenado el modelo, puede usarse para predecir la clase a la que pertenece una imagen que se le ingrese, dentro de las clases usadas en el entrenamiento. Para ello, la función desarrollada KNNprediction(), recibe como parámetros de entrada la imagen a clasificar y un vector con los valores de K según los que se hará la predicción. Si solo se desea usar un valor, el vector debe ser de una posición. La imagen ingresada se binariza y se le extraen sus características siguiendo el mismo proceso que las imágenes del conjunto de entrenamiento. Finalmente, estas características se ingresan al modelo a través de la función find_nearest(), que recibe como parámetros de entrada las características de la imagen y el valor de K para la clasificación. La función tiene tres salidas útiles: el resultado de la predicción, un vector con los K vecinos más cercanos y un vector con la distancia calculada entre las características ingresadas y los K vecinos más cercanos. Tanto la predicción como los K vecinos más cercanos retornados  tienen la forma de números enteros por lo que se emplea el diccionario creado al inicio para convertir los números a las clases reales de las imágenes.  Este procedimiento se repite para cada uno de los valores en el vector de K ingresado a KNNprediction(). </p>\n",
    "<p style=\"text-align:justify\">Por último, para realizar las matrices de confusión y la precisión del clasificador se utilizaron las funciones confusion_matrix() y accuracy_score() de la librería de Scikitlearn las cuales reciben como parámetros de entrada el conjunto de vectores correspondientes a los valores predichos y los valores verdaderos y retornan la matriz de confusión y la precisión del clasificador respectivamente. </p>\n",
    "<h3>- Análisis de los resultados:</h3>\n",
    "<p style=\"text-align:justify\">Se puede observar que cuando se realiza la clasificación con K igual a uno, los resultados tienden a ser erróneos. Esto era algo de esperarse ya que las características de la imagen pueden hacerla muy similar a un caso aislado de una clase a la que no pertenece. Clasificar a una imagen utilizando solo el vecino más cercano no resulta suficiente. Cuando se usa un K igual a tres, los resultados cambian drásticamente y dan una precisión del 100 por ciento con el conjunto de prueba, mostrando la mejora obtenida al contemplar la vecindad de las características de la imagen en lugar de solo el vecino más cercano. El uso de un K mucho mayor, por ejemplo un K igual a nueve, arroja la misma precisión. Sin embargo, podría haberse dado el caso que un K mayor resultara en una precisión menor, ya que el conjunto de entrenamiento no contiene el mismo número de imágenes para cada clase además de que el total de imágenes del conjunto de entrenamiento no está demasiado alejado de un K mayor que nueve. Adicional a lo anterior, es necesario agregar que es posible comprobar lo mencionado anteriormente a través de los resultados obtenidos tanto en el caso de la precisión en la clasificación como en el caso de las gráficas de las matrices de confusión. En ambos, se puede observar la diferencia de considerar un vecino cercano frente a usar más de uno. Finalmente, en cada una de las gráficas de las matrices de confusión es posible observar el porcentaje de correspondencia entre los valores de verdad frente a los valores predichos, siendo la diagonal principal de la matriz, el porcentaje de predicciones correctas para cada una de las clases. A partir de esta matriz, se puede observar que para el caso de 1 solo vecino los porcentajes de predicciones correctas resultan incluso del 0% para alguna de las clases a diferencia de considerar 3 y 9 vecinos cercanos donde los porcentajes son del 100% para cada una de las clases consideradas.</p>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 2",
   "language": "python",
   "name": "python2"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 2
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython2",
   "version": "2.7.12"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 0
}