Seperating each character.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import os\n",
    "import os.path\n",
    "import cv2\n",
    "import glob\n",
    "import imutils\n",
    "import matplotlib.pyplot as plt"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "CAPTCHA_IMAGE_FOLDER = 'electoral-tagged'\n",
    "OUTPUT_FOLDER = 'seperate'\n",
    "\n",
    "captcha_image_files = glob.glob(os.path.join(CAPTCHA_IMAGE_FOLDER, \"*\"))\n",
    "counts = {}"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Getting the first image of our dataset"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "captcha_image_file = captcha_image_files[0]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "19226.png\n"
     ]
    }
   ],
   "source": [
    "filename = os.path.basename(captcha_image_file)\n",
    "print(filename)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Storing the correct text of our captcha"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "19226\n"
     ]
    }
   ],
   "source": [
    "captcha_correct_text = os.path.splitext(filename)[0]\n",
    "print(captcha_correct_text)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Resizing, converting our captcha"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "image = cv2.imread(captcha_image_file)\n",
    "image = cv2.resize(image, (700,250))\n",
    "image = cv2.GaussianBlur(image,(1,9),0)\n",
    "image = cv2.pyrMeanShiftFiltering(image, sp=8, sr=16, maxLevel=1, termcrit=(cv2.TERM_CRITERIA_EPS+cv2.TERM_CRITERIA_MAX_ITER, 3, 1))\n",
    "gray = cv2.cvtColor(image, cv2.COLOR_BGR2GRAY)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## This is our captcha"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.imshow(gray, cmap='gray')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "thresh = cv2.threshold(gray, 0, 255, cv2.THRESH_BINARY | cv2.THRESH_OTSU)[1]\n",
    "thresh = cv2.erode(thresh, None, iterations = 3)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## After thresholding it looks like this"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.imshow(thresh, cmap='gray')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Finding contours in our captcha"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": [
    "contours = cv2.findContours(thresh.copy(), cv2.RETR_LIST, cv2.CHAIN_APPROX_SIMPLE)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [],
   "source": [
    "contours = contours[0] if imutils.is_cv2() else contours[1]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[[255, 120,   0],\n",
       "        [255, 120,   0],\n",
       "        [255, 120,   0],\n",
       "        ...,\n",
       "        [255, 120,   0],\n",
       "        [255, 120,   0],\n",
       "        [255, 120,   0]],\n",
       "\n",
       "       [[255, 120,   0],\n",
       "        [255, 120,   0],\n",
       "        [255, 120,   0],\n",
       "        ...,\n",
       "        [255, 120,   0],\n",
       "        [255, 120,   0],\n",
       "        [255, 120,   0]],\n",
       "\n",
       "       [[255, 120,   0],\n",
       "        [255, 120,   0],\n",
       "        [194, 194, 194],\n",
       "        ...,\n",
       "        [194, 194, 194],\n",
       "        [255, 120,   0],\n",
       "        [255, 120,   0]],\n",
       "\n",
       "       ...,\n",
       "\n",
       "       [[255, 120,   0],\n",
       "        [255, 120,   0],\n",
       "        [194, 194, 194],\n",
       "        ...,\n",
       "        [198, 198, 198],\n",
       "        [255, 120,   0],\n",
       "        [255, 120,   0]],\n",
       "\n",
       "       [[255, 120,   0],\n",
       "        [255, 120,   0],\n",
       "        [255, 120,   0],\n",
       "        ...,\n",
       "        [255, 120,   0],\n",
       "        [255, 120,   0],\n",
       "        [255, 120,   0]],\n",
       "\n",
       "       [[255, 120,   0],\n",
       "        [255, 120,   0],\n",
       "        [255, 120,   0],\n",
       "        ...,\n",
       "        [255, 120,   0],\n",
       "        [255, 120,   0],\n",
       "        [255, 120,   0]]], dtype=uint8)"
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "cv2.drawContours(image, contours, -1, (255,120,0), thickness = 2)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## These are the contours found"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "As we can see, there are some unwanted contours, we will remove them in the next step"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.imshow(image, cmap='gray')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [],
   "source": [
    "letter_image_regions = []"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "8\n"
     ]
    }
   ],
   "source": [
    "print(len(contours))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [],
   "source": [
    "for contour in contours:\n",
    "    # Get the rectangle that contains the contour\n",
    "    (x, y, w, h) = cv2.boundingRect(contour)\n",
    "        \n",
    "    letter_image_regions.append((x, y, w, h))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "8"
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(letter_image_regions)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [],
   "source": [
    "letter_image_regions = sorted(letter_image_regions, key = lambda x: x[0])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[(0, 0, 700, 250),\n",
       " (87, 61, 41, 69),\n",
       " (146, 55, 47, 69),\n",
       " (165, 70, 9, 14),\n",
       " (238, 61, 41, 69),\n",
       " (367, 96, 42, 70),\n",
       " (485, 31, 47, 69),\n",
       " (505, 71, 8, 13)]"
      ]
     },
     "execution_count": 19,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "letter_image_regions"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Removing unwated contours (those which start near origin | those whose width/height are less than 20)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[(87, 61, 41, 69), (146, 55, 47, 69), (238, 61, 41, 69), (367, 96, 42, 70), (485, 31, 47, 69)]\n"
     ]
    }
   ],
   "source": [
    "\n",
    "res = [i for i in letter_image_regions if min(i[1],i[2],i[0],i[3]) > 20]\n",
    "\n",
    "letter_image_regions = res\n",
    "\n",
    "print(letter_image_regions)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "5"
      ]
     },
     "execution_count": 21,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(letter_image_regions)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## These are the boundingRectangle of our final contours"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "((87, 61), (128, 130))\n",
      "((146, 55), (193, 124))\n",
      "((238, 61), (279, 130))\n",
      "((367, 96), (409, 166))\n",
      "((485, 31), (532, 100))\n"
     ]
    }
   ],
   "source": [
    "\n",
    "col = 20\n",
    "for i in letter_image_regions :\n",
    "    e = i[0]\n",
    "    f = i[1]\n",
    "    g = i[2]\n",
    "    h = i[3]\n",
    "    img = cv2.rectangle(image,(e,f),(e+g,f+h),(255,col,0),3)\n",
    "    col = col + 50\n",
    "    print((e,f),(e+g,f+h))\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.imshow(img, cmap='gray')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Extracting each letter using these boundingRectangles and saving them letterwise"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAKwAAAD8CAYAAADqv08vAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADl0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uIDIuMi4zLCBodHRwOi8vbWF0cGxvdGxpYi5vcmcvIxREBQAAGYhJREFUeJztnVusXVd1hv+BnfvNcWzHxjYxEREBVSK0RxQpVZVCqWhAJA9QkaIKtZHyQqugIpHQPrSVWim8AJVaIVlA60qUS7moUYRoo5Co6kuIAykluCEhSojx5cSJTUICCQ6jD3utnc/OGt5znb32Pnsej0+yztzrrMtc+0zP8c8xxxzT3F1JUguvWu0KJEkfssEmVZENNqmKbLBJVWSDTaoiG2xSFdlgk6qYqsGa2TvN7CEze8TMbh2qUkkSYSudODCzdZJ+KOkdkvZLuk/SDe7+g+GqlyQnsn6Ka98i6RF3f1SSzOyLkq6TFDbYTZs2+a5du6Z4ZLJWeeyxx3TkyBGbdN40DXa7pCfweb+k3zzVBbt27dLevXuneORwRJbFbOJ3tnDwXWqsvyQtLS0VnTeNhu36Zl7RCszsJjPba2Z7n3zyySkelyTTNdj9knbi8w5JB04+yd13u/uSuy9t3rx5isclxN3H/04npmmw90m6wsxea2ZnSnq/pNuHqVaSdLNiDevux83sTyX9h6R1kj7n7g8OVrMk6WCaQZfc/RuSvjFQXeZKrYOTlr71XwsDMylnupLKyAabVMVUkmCtEI20efyFF17oPP6rX/1q4v15zqte1a+P4Plnn332uByZ9RJzP8mzsMiSIXvYpCqywSZVUb0kiEa/NOE0yS+99NIJP08uP//88533JrxfyRRvVMe+165f3/3nolTgOevWrRuXzzvvvM77d7HI09bZwyZVkQ02qYrqJQFN//Hjx8flF198sfOc9jjN3i9+8YtxmZKA96MMiMolpr/keARN/5lnnjkun3POOeMyJQHLzz33XOd9zjrrrFccW2Syh02qIhtsUhVzlwStGZxmxEkTzvLPf/7zcZkyoOscmkh6CZ555plx+Ze//GXnOZQBvDclRGTuWS6ZROBI/9xzzx2XOeqnPIikwvnnn3/KerIuvAdZhHiE7GGTqsgGm1RFNV4CmmGafpp2mnAe58j/2LFjkqSf/exnnb+nJOBzaA5pUvkcHo/gfUoc9DTPF1988bh80UUXjctnnHHGuEzPAFd40GvCcisFeF0kCRaB7GGTqsgGm1RFNZKAZpum+tlnnx2X6RmgyaeZb4+30uDk3//0pz8dl+kZ4DN5bx6nJOC1JBpdRyNwju7pkVheXh6XN23aNC5THrA+W7Zs6XxuO3FAScBnLhrZwyZVkQ02qYpqJAFNLE0/j0+SAZL09NNPSzpRShw9enRcplSgDIliFuglYLlkJUIEnfisO+u2YcOGcZnfAU07vSYRrYRopcHJz6Q84WTFak0iTOxhzexzZrZsZt/HsY1mdqeZPdz8vPhU90iSoSjpYf9Z0j9I+hccu1XSXe5+W5Nm81ZJtwxfvZeZNNUqndjDRb1tew6vY2/IXon+SvZiPCe6T9/g7MiHGwWXE773BRdc0FmfqA7t1C99r5wO5iCO9V2tgdnEHtbd/0vS0ycdvk7Snqa8R9L1A9crSTpZ6aDrUnc/KEnNz26fiTIZXDIsMx90uftuSbslaWlpqVfmMpr+aNATTavSn0p50JpJmloOcjiYieoSmUxKBdYl8snSZHOwVLLs/KmnnhqXOWXLOkTLy1n/liiijIHdUeA6p4lnzUp72MNmtk2Smp/LE85PkkFYaYO9XdIHm/IHJf37MNVJklMzURKY2RckXSNpk5ntl/RXkm6T9GUzu1HSjyW9bxaVo9mhDKB5bv2q0omjZUoC0ppV+h1pIlmO8tmWLBHntVFEGb0QlBCRxyAa9VMeRFASdJUjSUK5weNRRhp+r7NgYoN19xuCX7194LokyURyajapioWemqW5pZmks54mlvKgK0hZ6naU06RF8oD3KxnF03xzSpXvwfpy/RUlT+RtYJnPLclc0zWVGsmNSJ7wu5mnJMgeNqmKbLBJVSy0JKDZibKw0FTzOCcLOO/djoZ5jCaNEUmUEtOkaKckoPeCI3OOxo8cOdJ5Do8Tvivvz+9sUv1LostYF5p+/g1mTfawSVVkg02qYqElQWSmorC8aM6epr2d+6ckYEaVSy65ZFyOlltHTnaeQ+8FQ/5o+ilFOLqnFGId+H1wTRfjLBiYzjqwzpMyzkTeBcoAriObJli9L9nDJlWRDTapioWWBJHp4vGSmACawzYULpIBGzdu7LwuGiFHSdTo/GcIIp9LaOJJlKiOUiEKa2QMQ5QdZhKc0IhWdFCSzJrsYZOqyAabVMXcJcFKlwTTjEWj9yjtOSPiW+8ATT89BizT9NOkRvlYeX60YDDKxhJ9L4w3uPDCC8dlLjeKFgpGKzPoSWglTRSiSG8AJ0CivLizJnvYpCqywSZVsdBeAo5QaW5pzqOFcRzR0pS21zJmgDKAx6OcqdFIn/Acmk+uRKBnIFrdwFBKmu2SOpDIw9ASLVikJNm6deu4TO8F32/WZA+bVEU22KQqFloS0HRxpE9HeTT3TxPHSYQ27wAlBmUAZQjLfbeNJ7w/60UvARPSkUOHDo3LjEOIvAQRHMmzTKnQQvnTtU/vyfXdvn37xOcPRUkyuJ1mdreZ7TOzB83s5uZ4JoRL5k5Jt3Fc0kfc/Q2S3irpQ2b2Rr2cEO4KSXc1n5NkppQs8z4oqc2j9ayZ7ZO0XaOEcNc0p+2RdI9mmMGQZo9eAsoAjuppvnhtKwU4yo5WHAxFtDAvCvlj+dJLLx2XDx48OC5T5vTd9YUyoA195PfF+tIzQRmybdu2Xs8cil7CzMx2SXqzpHvVIyFckgxFcYM1s/MlfVXSh939lUo9vi6zFyaDUeQlMLMzNGqsn3f3rzWHD5vZNnc/eKqEcNNkLyQ04TSZNKtROqOuEETKBEqCkl1epqEkPoHvRBlAKUSvyTQLJNu4An6n9I7QI1OSj2HWlHgJTNJnJe1z90/gV5kQLpk7JT3s1ZL+SNL/mtkDzbG/0JwSwiUJKfES/LekyOasSkI4mnCWo1RBK3X6z9rUsV5RbEBJ6p9pdnFp35HfF+UUYzK6rpv2+X3JqdmkKrLBJlWx0LEEEX1Hq13r5udpxvpCeRAtpoy2TyI8Hu1B23Uuf0/ZEoVyLpSXIEkWiWywSVVUKQkicxht3TNPkzU0Ua6FkneKchf0kUMlHpb0EiRJQDbYpCqqlASkZO6/j8laZPkQ1S3K2rhSIhmwCN9N9rBJVWSDTaqiekkQsVIvwTTehaE8E9HCv8hxP/RIfpoFl7NmcWuWJB1kg02qYs1KArIa8mAaGN4XyQMyhOOe92DcQbS6Y7VkQ/awSVWcFj3sEMxi+rEkaJqbtpXsd9uXtqdkhBiDxvsmnZs12cMmVZENNqmKlARzIEpSxxTsjKbifrEsUzawzGtL4ICpTabHgRY3wuvKmiPNdzM5UrLM+2wz+7aZ/U+TDO5vmuOvNbN7m2RwXzKzfvlykmQFlEiCFyS9zd3fJOkqSe80s7dK+rikTzbJ4I5KunF21UySESXLvF1Su4vYGc0/l/Q2SX/YHN8j6a8lfXr4KtZJJAMiE8+N2pimnbu/0HsQTdmW1Ic7w7RZXugNYAp9ygMmymvz7M6bokGXma1rkmgsS7pT0o8kHXP3drXafo0yGibJTClqsO7+krtfJWmHpLdIekPXaV3XZjK4ZEh6eQnc/ZiZ3aNRYuMNZra+6WV3SDoQXDNIMrihaE3jPKdauRMM92WlDGDOVp7PTeA4ccBzSiQBzTwTz3VJAsoA7sBDj0HJ/rmzoMRLsNnMNjTlcyT9rqR9ku6W9N7mtEwGl8yFkh52m6Q9ZrZOowb+ZXe/w8x+IOmLZva3kr6rUYbDJJkpJV6C72mUdfvk449qpGeTCUSmnDKApp9eAp7DMj0MvH9ElJe2lQQ0/UxTTynBHXtKcurOgpyaTaoiG2xSFadtLME8g7OjCQLKA8oAuv8oA+hhKJEBfEemgac5b1PPc9O6aB9e3o+ehpIctkORPWxSFdlgk6o4bSVB32RqfWUDz6ckYGxAFEZIeUDvAcu8Z1S3KLds1yQBvQSMNaBUYPwAZUVfppFg2cMmVZENNqmK01YSzBp6A6KFhJQB9AxQEnCv10haMMU7ZQw3oqMJ37x587jcmvmtW7e+4ph0oseAMQb0OkRegll4X7KHTaoiG2xSFaedJOgyUyVz4SXnRDux0JQfPXp0XI48AEeOHBmXKS2iTd6ietK5T9NOL0HrBaB8oJeA8QO8B+/N8qzJHjapimywSVWcFpJgiNFqdI9IKkQjenoJoviBSDZQHnCVAevAuX+O3qNVBK0koAzo8iKcfB29BAu14iBJFolssElVrFlJ0NeET7ougqaZ4X8sM5UQ0xNRKtD0Ux5QWjAXQbQfLE1/12JD6UTT3p4fSQLej5MPlBuMWehL39UK2cMmVZENNqmKNSsJ+hBtQleynT2d+TTxHNEzHoCmP8pFwGujNPGUATTVlAp0+nO0z5DBVgrw97wuWlnQrlSYN8U9bJOu6LtmdkfzObMXJnOnjyS4WaMEGi2ZvTCZO0WSwMx2SHqXpL+T9Oc2spsLkb1w6O3no3Pp8I8WBtKU0/lP2cCQQnoMonvS9DOfQBQzwHNo+mnmac5bKcAJB0oMTgrw3tN4Bkhfr0xpD/spSR+V1Iq3S1SYvTCTwSVDUpJb692Slt39fh7uOLXzv4q773b3JXdfon8vSVZCiSS4WtJ7zOxaSWdLulCjHrcoe+GsGSI0sEQ+lIzuo9gAygDeJ5ITHOlHKYFo1mm2GQK4ffvLRo+dBaVCe58o1oByI5IB88wEObGHdfePufsOd98l6f2SvuXuH1BmL0xWgWkmDm7RaAD2iEaaNrMXJjOnb0LjeyTd05Srzl5YYsZoyjnSL1kpsLy8PC5THkRbHTFmgDCfQJQ7gJKA3gBmIdyyZcu43JWWiNIjyjkQeWTmSU7NJlWRDTapitMulqDPiDaKGaAMOHz48LhM0x9JAsYb0NtAOFnAkTllAEf19ABs27ZtXKYkoAzggsNWEjD8sIR5egZI9rBJVWSDTaritJAEKzVfNN8c3dPEs0x5QEnAa0uyDlIScIKAZpsygBMEr371q8dlegzoYegqU2IsMtnDJlWRDTapijUrCSaFCdJpz7l7Ho8SC3MFASUBj0cTBHwWiZz1HNFHI33Kgx07dozLNP2TNj+mDJknuQgxWdNkg02qYk1JgkgGdG0XxIkAjtxZplmPwv+i8ykDuJCQEwEM3eMiQL4HwwU5EcBytMCQI38uIKQkGGrlwLzIHjapijXVw0Z0pWmP9nZl7xZNtUY+1pJeleVoOpRrp3bu3DkuM+KKAy0OwNirRonhWIdpdoNZDbKHTaoiG2xSFdVLgpJp165l1lyGTRNPXyqlBIOzGWXF6dsoSwtNPE0w/aRRoDZNf5TKnYMu+lt5PFoCv9rMapl3kiwE2WCTqqheEkTmjX5TmvBWCkQB1ocOHRqXo6Bt+l6jbeBp4ikDOKKP1mjxOEf6UfYWXkvPA6dbo3yy7fe3WjKh73NLUxU9JulZSS9JOu7uS2a2UdKXJO2S9JikP3D3o9E9kmQI+kiC33H3q9x9qfl8q6S7mmRwdzWfk2SmTCMJrpN0TVPeo9Hy71umrM9UcA0WM7LQhLdSgKP+n/zkJ+Py448/3nk/TgqQaJRLRz1H6yxzepVSgRMEnAigKec59EJEO7p0yYAaKe1hXdJ/mtn9ZnZTc+xSdz8oSc3PLeHVSTIQpT3s1e5+wMy2SLrTzP6v9AFNA79Jkl7zmtesoIpJ8jJFDdbdDzQ/l83s6xplfDlsZtvc/aCZbZO0HFy7W9JuSVpaWhpkbTBNNb0BNNuRl6A9Tm/AE088MS5TKnAigCP9KDYg8gxEI3qW+Z+5KzOLdKIppzeA50REMmA15ME0zyxJt3memV3QliX9nqTvS7pdoyRwUiaDS+ZESQ97qaSvN/8r1kv6V3f/ppndJ+nLZnajpB9Let/sqpkkIyY22Cbp25s6jj8l6e2zqFQXDMKmQ5/H6dxn2B8nCdpQwijR28GDB8dlmnXKDY70o3BBjvqj5dnRbi08h+Zzmj1da/YMkJyaTaoiG2xSFdXEEtBss8wwQZp5jva7ErZF8QP0KNAbEaVO5/Eo8p8TBLyWcQKcCCgZ9dODEU0KLEI+16HJHjapimywSVXMXRK0ZqrERNGk0RvAVQGUBFxMSM8AJUHrBaBHgZMMPM5ROeUBTTCd//QebN26dVyOwgin2ait5Py1IgNI9rBJVWSDTapiob0E9AZw9E5vAGUAwwQpCXi8Nf+UCQxFLEncRs8AzT1lACcFWKYHgPKA4YhDsUhegqGenz1sUhXZYJOqWGhJwJUC9BJQHnDunxMAlAE8vz2HngHKCkoCLjDkqDza5SXa7p3eBp5PM8n3K5k4WCTmKTeyh02qIhtsUhULLQk4yqUJp6mmOT9w4EDnccYVtPfh5AO9Abw3y1EaoogoNRDfiZ4MxhIwziG659BE945WNuRes0lSQDbYpCqqkQQcvdOUcnKBqwz2798/LtPkt+dEpp/PpAxgLEHJHrT0JDDsMAoFZB2nSeM+hKlmHRkCyUyNURbGWZM9bFIV2WCTqihNBrdB0mck/ZpGWWD+RNJDmmMyOJpnhuXRnEe7uzD2oJUW0XU0qZQhvB89FpQHlCGUFpQqhPkKpjHlQ68yiORMlHIpkhCzoLSH/XtJ33T3KzVaQbtPmQwuWQVKEmlcKOm3JX1Wktz9RXc/plEyuD3NaXskXT+rSiZJS4kkuFzSk5L+yczeJOl+STfrpGRwTd6tQYmi/Lv2LJBOnAzokgHSy3P2vDflBuUBj3OunxMRNLuve93rxmXKBsLz6Yjvm12wb+qhlcoD5kjg+xGGWy6CJFgv6dclfdrd3yzpOfUw/2Z2k5ntNbO90QxOkpRS0mD3S9rv7vc2n7+iUQM+3CSB06RkcO6+5O5L/N+aJCuhJFXRITN7wsxe7+4PaZSe6AfNvw9Kuk0zSgYXRfzTxHNUHIUj0vy35/MYzSWd9nw+YxP6OvaZ1TBKNxSN7meRdXDSVvP0DBCGT/L59BjMmtKZrj+T9HkzO1PSo5L+WKPeOZPBJXOlND/sA5KWOn41t2RwSSIteCxBCX1NY3s+zSJNNqUCoTzgCoZoooH3jIikRRTKyGdNMuununbS7oNcLREdpzyIpFvENJMbOTWbVEU22KQqqpQEjCWgaWTIG9f502Tx2pYuL4J0osmOnknzTUlQskIhMuslXoJZ5hygR4RlTszQRUmPTF9yc+RkTZMNNqmKKiUBkwVztMqFfIyO74pJiLwBJMohQFMe5SgoMdORlyAyk6udbiiCsQSzJnvYpCrm3sP26SU4iIqimZiOnblih6jHUOdMQ4m/tcRa9IFTrUxYx4EWLVvXQHZWZA+bVEU22KQqFnrQRV/qZZddNi5HftMrr7xyXI4ii/pQMp05z4FQ32f19XG2UIrxe+fULAe+fZnpXrNJskhkg02qYqElAc0O12hxFHv55ZePyzQ1URDyJEqmQheBvrvwrPTeUYQWl6hn5pckCcgGm1TFQksCwrVQdGbTZPUNau5iUac/VwvKAC7h5vGUBEkSkA02qYpqJAFH/XRsc1M4egzIpNFy33Twi0af9Vp970f4N5hm4mAaSnJrvd7MHsC/Z8zsw2a20czuNLOHm58XT7pXkkzLxAbr7g+5+1XufpWk35D0vKSvK7MXJqtAX0nwdkk/cvfHzew6Sdc0x/dIukfSLcNVLYYegyiTCrPAdIXfUVZMsyapdvg9dLFoXpO+g673S/pCUz4he6GkwbMXJsnJFDfYJk3ReyT9W58HZPbCZEj6SILfl/Qdd2/3bT9sZtua3LCnzF4oabckLS0tzW1CfpKpI9w8LVls+kiCG/SyHJCk2zXKWijNKHthkpxMUYM1s3MlvUPS13D4NknvMLOHm9/dNnz1kuRESrMXPi/pkpOOPaXMXpjMmZyaTaoiG2xSFdlgk6rIBptURTbYpCpsnovrzOxJjfb5OjLp3DXCJuW7lnKZu0/cF2uuDVaSzGyvu3dt8LHmyHcdnpQESVVkg02qYjUa7O5VeOZqke86MHPXsEkyDSkJkqqYa4M1s3ea2UNm9oiZrZk1YGa208zuNrN9Zvagmd3cHF+zCzXNbJ2ZfdfM7mg+v9bM7m3e9UtNwP/gzK3Bmtk6Sf+oUSD4GyXdYGZvnNfzZ8xxSR9x9zdIequkDzXvtpYXat4saR8+f1zSJ5t3PSrpxlk8dJ497FskPeLuj7r7i5K+KOm6OT5/Zrj7QXf/TlN+VqM/5HaN3m9Pc9oeSdevTg2Hxcx2SHqXpM80n03S2yR9pTllZu86zwa7XdIT+Ly/ObamMLNdkt4s6V6t3YWan5L0UUntcuRLJB1z93bLyZn9befZYLvWC68pF4WZnS/pq5I+7O7PTDq/Rszs3ZKW3f1+Hu44dSZ/23mmKtovaSc+75B0YI7PnylmdoZGjfXz7t4uJSpaqFkZV0t6j5ldK+lsSRdq1ONuMLP1TS87s7/tPHvY+yRd0Ywmz9Qox8Htc3z+zGg03Gcl7XP3T+BXa26hprt/zN13uPsujf6G33L3D0i6W9J7m9Nm967uPrd/kq6V9ENJP5L0l/N89ozf67c0MoHfk/RA8+9ajbTdXZIebn5uXO26Dvze10i6oylfLunbkh7RKHfFWbN4Zs50JVWRM11JVWSDTaoiG2xSFdlgk6rIBptURTbYpCqywSZVkQ02qYr/B4rQBfB26i0IAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "for letter_bounding_box, letter_text in zip(letter_image_regions, captcha_correct_text):\n",
    "    # Grab the coordinates of the letter in the image\n",
    "    x, y, w, h = letter_bounding_box\n",
    "\n",
    "    # Extract the letter from the original image with a 2-pixel margin around the edge\n",
    "    letter_image = gray[y - 2:y + h + 2, x - 2:x + w + 2]\n",
    "    \n",
    "    plt.imshow(letter_image, cmap='gray')\n",
    "    plt.show()\n",
    "    \n",
    "    \n",
    "    save_path = os.path.join(OUTPUT_FOLDER, letter_text)\n",
    "\n",
    "    if not os.path.exists(save_path):\n",
    "        os.makedirs(save_path)\n",
    "        \n",
    "    count = counts.get(letter_text, 1)\n",
    "    p = os.path.join(save_path, \"{}.png\".format(str(count).zfill(6)))\n",
    "    cv2.imwrite(p, letter_image)\n",
    "    \n",
    "    counts[letter_text] = count + 1\n",
    "\n",
    "    "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "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": "2.7.16"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}