ocr-test.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "try:\n",
    "    import Image\n",
    "except:\n",
    "    from PIL import Image\n",
    "\n",
    "import pytesseract\n",
    "import os.path\n",
    "from pytesseract import image_to_string\n",
    "%matplotlib inline\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 99,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "from group_nodes_2_text import *"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "/Users/celder/git/RC-Tree/images/Tree2_Circular.png\n"
     ]
    }
   ],
   "source": [
    "#file = 'Tree1.png'\n",
    "#file = 'tree_image_ref.png'\n",
    "#file2 ='test-european.jpg'\n",
    "#file = 'tree_image_ref_right.png'\n",
    "file = 'images/Tree2_Circular.png'\n",
    "script_dir = os.path.dirname(os.path.abspath('__file__'))\n",
    "print os.path.join(script_dir, file)\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "PNG (1187, 1190) RGBA\n"
     ]
    }
   ],
   "source": [
    "\n",
    "image = Image.open(os.path.join(script_dir, file))\n",
    "print image.format, image.size, image.mode\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Passer_monlanus\n",
      "\n",
      " \n",
      "\n",
      "e\n",
      ".m\n",
      "d\n",
      "n\n",
      "k a\n",
      "o\n",
      "n s\n",
      "e n .00\n",
      "m a s Q .8». 6%\n",
      "m Wm“ My ya la 29 Leo a,»\n",
      "wig 405% $90 $2 3%\n",
      ".m I m%.% \\ a: W 96\n",
      "md mnwﬂwwwmshéy 0% Woo\n",
      "mi 5% 4 vkwsvés 3,6,6\n",
      "M 362663430359 am\n",
      "3:05.: : mauvawﬁmx maw\\=a%\n",
      "kc 03ml 110m? 2.36m ix\n",
      "ncEEEImSE nm_mnm:o_%m\\===m\n",
      ")mu<n_=m|=5:mn3=m ogazmnEmmenaEmim\n",
      ".m\n",
      "oéoamax‘maoaa. oﬂoamniﬂgnoaamm N\n",
      "GEE mSES J I\n",
      "303%,: NkaQI .‘EQ MEESQ\n",
      "34.6 mm mxme\n",
      "760%? . 3am:\n",
      "056% m Sees «an»;\n",
      ":m\\ o . uh»?\n",
      ".328 oxsoo 5; ﬁe % m ovxwo/qemﬁz\n",
      "39 ac 1/94 a,» a, we a 5» a& + am\n",
      "% doavsqwmm a— \\oxsoe $3\n",
      "a» 909 W? W W ,m 9 @00¢ \\eQ/ p.36» axe?\n",
      "\\ 8 0 I b\n",
      "2% +v Wu» \"we AW 5%. pvév 0/ +@ \\mwmx\n",
      "£0» 0.9 «A, ,o{ M W; m ﬂaw \\e/ Am 59¢ &@\n",
      "o\n",
      "A: oimuh m WN/ﬁwwoa aoav ”we\n",
      "mhwmmn wwwsuwaawwq\n",
      "9% J .1 m M .W e M .w .m\\ @w.\n",
      "m w x, m g no 4» we 40\n",
      ".0 i m. a, s, , ﬂ 0 ﬁx\n",
      "r m. m m a M w e +59\n",
      "m & m M A no! mx\n",
      "S R B H\n"
     ]
    }
   ],
   "source": [
    "print(pytesseract.image_to_string(image))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "image = Image.open(os.path.join(script_dir, file2))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "#print(pytesseract.image_to_string(image, lang='fra'))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 277,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Phrase orig: Fufica atra\n",
      "\n",
      "Phrase orig with configs: Fufica atra\n",
      "\n",
      "Phrase orig with verbose configs: Fufica atra\n",
      "\n",
      "Phrase grayscale: Fuﬁca atria\n",
      "\n",
      "\n",
      "Phrase grayscale with configs: Fuﬁca arra\n",
      "\n",
      "Phrase grayscale with verbose configs: Fuﬁca atria\n",
      "\n",
      "Phrase grayscale with extra verbose configs: Fuﬁca atria\n",
      "\n",
      "\n",
      "Phrase thresholded: Fuh'ca afra\n",
      "\n",
      "Phrase thresholded with configs: Fuh'ca ax‘ra\n",
      "\n",
      "Phrase thresholded with verbose configs: Fuh'ca afra\n",
      "\n"
     ]
    }
   ],
   "source": [
    "from PIL import Image\n",
    "import pytesseract\n",
    "import os.path\n",
    "script_dir = os.path.dirname(os.path.abspath('__file__'))\n",
    "from pytesseract import image_to_string\n",
    "import cv2\n",
    "import numpy\n",
    "\n",
    "# file = 'images/Tree2.png'\n",
    "# image = Image.open(os.path.join(script_dir, file))\n",
    "# print(\"Tree1\" + pytesseract.image_to_string(image) + \"\\n\")\n",
    "\n",
    "file_word = 'images/fa_phrase.png'\n",
    "# file_word = 'images/rost_phrase.png'\n",
    "# file_word = 'images/fc_phrase.png'\n",
    "# image_word = Image.open(os.path.join(script_dir, file_word))\n",
    "# image_word = cv2.imread(file_word)\n",
    "# print(\"Troodon word: \" + pytesseract.image_to_string(image_word) + \"\\n\")\n",
    "# image_word.cvtColor(image_word, image_word_gray, CV_BGR2GRAY)\n",
    "\n",
    "im=cv2.imread(file_word)\n",
    "# cv2.imshow('Original',im)\n",
    "im_image = Image.fromarray(im)\n",
    "\n",
    "print(\"Phrase orig: \" + pytesseract.image_to_string(im_image) + \"\\n\")\n",
    "cv2.imwrite('images/phrase_orig.png', im)\n",
    "\n",
    "# print(\"Phrase orig with configs: \" + pytesseract.image_to_string(im_image, config=\"--user-words config/user-words.txt, --user-patterns config/user-patterns.txt, --tessedit_char_whitelist AaBbCcDdEeFfGgHhIiJjKkLlMmNnOoPpQqRrSsTtUuVvWwXxYyZz\") + \"\\n\")\n",
    "print(\"Phrase orig with configs: \" + pytesseract.image_to_string(im_image, config=\"--bazaar config/bazaar\") + \"\\n\")\n",
    "print(\"Phrase orig with verbose configs: \" + pytesseract.image_to_string(im_image, config=\"--user-words config/user-words.txt --user-patterns config/user-patterns.txt\") + \"\\n\")\n",
    "\n",
    "gray_im = cv2.cvtColor(im, cv2.COLOR_BGR2GRAY)\n",
    "# cv2.imshow('Gray',gray_im)\n",
    "gray_image = Image.fromarray(gray_im)\n",
    "print(\"Phrase grayscale: \" + pytesseract.image_to_string(gray_image) + \"\\n\")\n",
    "cv2.imwrite('images/phrase_gray.png', gray_im)\n",
    "\n",
    "print(\"\\nPhrase grayscale with configs: \" + pytesseract.image_to_string(gray_image, config=\"--bazaar config/bazaar\") + \"\\n\")\n",
    "print(\"Phrase grayscale with verbose configs: \" + pytesseract.image_to_string(gray_image, config=\"--user-words config/user-words.txt --user-patterns config/user-patterns.txt\") + \"\\n\")\n",
    "print(\"Phrase grayscale with extra verbose configs: \" + pytesseract.image_to_string(gray_image, config=\"--user-words config/user-words.txt --user-patterns config/user-patterns.txt --tessedit_char_whitelist AaBbCcDdEeFfGgHhIiJjKkLlMmNnOoPpQqRrSsTtUuVvWwXxYyZz_\") + \"\\n\\n\")\n",
    "\n",
    "thresh_im  = numpy.array(gray_im, numpy.uint8)\n",
    "thresh, thresh_im = cv2.threshold(gray_im, 125, 255, cv2.THRESH_BINARY)\n",
    "thresh_image = Image.fromarray(thresh_im)\n",
    "print(\"Phrase thresholded: \" + pytesseract.image_to_string(thresh_image) + \"\\n\")\n",
    "cv2.imwrite('images/phrase_thresh.png', thresh_im)\n",
    "\n",
    "print(\"Phrase thresholded with configs: \" + pytesseract.image_to_string(thresh_image, config=\"--bazaar config/bazaar\") + \"\\n\")\n",
    "print(\"Phrase thresholded with verbose configs: \" + pytesseract.image_to_string(thresh_image, config=\"--user-words config/user-words.txt --user-patterns config/user-patterns.txt\") + \"\\n\")\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "# import tesseract\n",
    "# ocr = tesseract.TessBaseAPI();\n",
    "# ocr.Init(\".\",\"eng\",tesseract.OEM_TESSERACT_ONLY)\n",
    "# ocr.SetVariable(\"tessedit_char_whitelist\", \"AaBbCcDdEeFfGgHhIiJjKkLlMmNnOoPpQqRrSsTtUuVvWwXxYyZz\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 139,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'Archoeopteryx Iithographica'"
      ]
     },
     "execution_count": 139,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# image to str as a function\n",
    "from PIL import Image\n",
    "import pytesseract\n",
    "# import os.path\n",
    "# script_dir = os.path.dirname(os.path.abspath('__file__'))\n",
    "from pytesseract import image_to_string\n",
    "import cv2\n",
    "import numpy\n",
    "\n",
    "def imgToStr(im): \n",
    "#     print(\"Phrase orig: \" + pytesseract.image_to_string(im_image) + \"\\n\")\n",
    "#     cv2.imwrite('images/phrase_orig.png', im)\n",
    "\n",
    "    # print(\"Phrase orig with configs: \" + pytesseract.image_to_string(im_image, config=\"--user-words config/user-words.txt, --user-patterns config/user-patterns.txt, --tessedit_char_whitelist AaBbCcDdEeFfGgHhIiJjKkLlMmNnOoPpQqRrSsTtUuVvWwXxYyZz\") + \"\\n\")\n",
    "#     print(\"Phrase orig with configs: \" + pytesseract.image_to_string(im_image, config=\"--bazaar config/bazaar\") + \"\\n\")\n",
    "#     print(\"Phrase orig with verbose configs: \" + pytesseract.image_to_string(im_image, config=\"--user-words config/user-words.txt --user-patterns config/user-patterns.txt\") + \"\\n\")\n",
    "#     print(\"Phrase orig with extra verbose configs: \" + pytesseract.image_to_string(im_image, config=\"--user-words config/user-words.txt --user-patterns config/user-patterns.txt --tessedit_char_whitelist AaBbCcDdEeFfGgHhIiJjKkLlMmNnOoPpQqRrSsTtUuVvWwXxYyZz_0123456789\") + \"\\n\\n\")\n",
    "\n",
    "    gray_im = cv2.cvtColor(im, cv2.COLOR_BGR2GRAY)\n",
    "    # cv2.imshow('Gray',gray_im)\n",
    "    gray_image = Image.fromarray(gray_im)\n",
    "#     print(\"\\nPhrase grayscale: \" + pytesseract.image_to_string(gray_image) + \"\\n\")\n",
    "#     cv2.imwrite('images/phrase_gray.png', gray_im)\n",
    "#     print(\"Phrase grayscale with minimal configs: \" + pytesseract.image_to_string(gray_image, config=\"--tessedit_char_whitelist AaBbCcDdEeFfGgHhIiJjKkLlMmNnOoPpQqRrSsTtUuVvWwXxYyZz_ --psm 7\") + \"\\n\")\n",
    "#     print(\"Phrase grayscale with configs: \" + pytesseract.image_to_string(gray_image, config=\"--bazaar config/bazaar\") + \"\\n\")\n",
    "#     print(\"Phrase grayscale with verbose configs: \" + pytesseract.image_to_string(gray_image, config=\"--user-words config/user-words.txt --user-patterns config/user-patterns.txt\") + \"\\n\")\n",
    "#     print(\"Phrase grayscale with extra verbose configs: \" + pytesseract.image_to_string(gray_image, config=\"--user-words config/user-words.txt --user-patterns config/user-patterns.txt --tessedit_char_whitelist AaBbCcDdEeFfGgHhIiJjKkLlMmNnOoPpQqRrSsTtUuVvWwXxYyZz_\") + \"\\n\\n\")\n",
    "\n",
    "#     thresh_im  = numpy.array(gray_im, numpy.uint8)\n",
    "#     thresh, thresh_im = cv2.threshold(gray_im, 125, 255, cv2.THRESH_BINARY)\n",
    "#     thresh_image = Image.fromarray(thresh_im)\n",
    "#     print(\"Phrase thresholded: \" + pytesseract.image_to_string(thresh_image) + \"\\n\")\n",
    "#     cv2.imwrite('images/phrase_thresh.png', thresh_im)\n",
    "\n",
    "#     print(\"Phrase thresholded with configs: \" + pytesseract.image_to_string(thresh_image, config=\"--bazaar config/bazaar\") + \"\\n\")\n",
    "#     print(\"Phrase thresholded with verbose configs: \" + pytesseract.image_to_string(thresh_image, config=\"--user-words config/user-words.txt --user-patterns config/user-patterns.txt\") + \"\\n\")\n",
    "    \n",
    "#     str = pytesseract.image_to_string(im, config=\"--user-words config/user-words.txt --user-patterns config/user-patterns.txt --tessedit_char_whitelist AaBbCcDdEeFfGgHhIiJjKkLlMmNnOoPpQqRrSsTtUuVvWwXxYyZz_0123456789\")\n",
    "    str = pytesseract.image_to_string(gray_image, config=\"--user-words config/user-words.txt --user-patterns config/user-patterns.txt --tessedit_char_whitelist AaBbCcDdEeFfGgHhIiJjKkLlMmNnOoPpQqRrSsTtUuVvWwXxYyZz_0123456789\")\n",
    "\n",
    "    return str\n",
    "\n",
    "    \n",
    "\n",
    "# file_word = 'images/sm_phrase.png'\n",
    "# file_word = 'images/rb_phrase.png'\n",
    "# file_word = 'images/mm_phrase.png'\n",
    "# file_word = 'images/bj_phrase.png'\n",
    "# file_word = 'images/gp_phrase.png'\n",
    "# file_word = 'images/fa_phrase.png'\n",
    "file_word = 'images/al_phrase.png'\n",
    "# file_word = 'images/tm_phrase.png'\n",
    "\n",
    "im=cv2.imread(file_word)\n",
    "cv2.imwrite('images/phrase_orig.png', im)\n",
    "im_image = Image.fromarray(im)\n",
    "\n",
    "imgToStr(im_image)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 278,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "def labels_for_tree():\n",
    "    imfile = \"images/tree_image_ref.png\"\n",
    "    just_tree = cv2.imread(imfile) #use k=9 to get only tips, k=4 gets some of the labels but is not as clean\n",
    "    result = label_rectangles(just_tree)    \n",
    "    labels = [(pos, get_label(Image.fromarray(just_tree), pos) ) for pos in result[1] ]\n",
    "    return labels\n",
    "#     nodes = le_main(imfile, 1)\n",
    "#     associations = []\n",
    "#     for label in labels:\n",
    "#         r,lname = label\n",
    "#         x,y,w,h = r\n",
    "#         # get distances to all nodes\n",
    "#         distances = set()\n",
    "#         for node in nodes:\n",
    "#             x1,y1 = node\n",
    "#             d = distance.euclidean((x,y),node)\n",
    "#             distances.add((d,node))\n",
    "#         #get the node that is the minimum distance and name it\n",
    "#         associations.append((min(distances)[1],(x,y),lname))\n",
    "#     tree = result[0].copy()\n",
    "#     map(print_leaf, associations)\n",
    "#     show(tree, False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 265,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "def get_label(img, pos):\n",
    "    x,y,w,h = pos\n",
    "    box = (x, y, x + w, y + h)\n",
    "    ocr_ready = img.crop(box)\n",
    "#     return ocr_ready\n",
    "    return imgToStr(ocr_ready)\n",
    "\n",
    "    "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Asdsa\n"
     ]
    }
   ],
   "source": [
    "print \"Asdsa\""
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 261,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAASwAAABFCAIAAAB/k0wEAABR2klEQVR4nO39V5dt13XnCc7ltt/7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\nE+EZZ3zOnInwjDM+Z/7/db18VXWfbAAAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<PIL.Image._ImageCrop image mode=RGB size=300x69 at 0x10D3F8750>"
      ]
     },
     "execution_count": 261,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "get_label(Image.fromarray(im), (10,100,69,300))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 235,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAOYAAAA7CAIAAADgqGEdAAA9UklEQVR4nO29V5MlR5bnd1yFFlfL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\nxv8P+tczP+tH5BQAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<PIL.Image._ImageCrop image mode=RGB size=230x59 at 0x10D44C110>"
      ]
     },
     "execution_count": 235,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "image = Image.fromarray(im)\n",
    "image.crop((70,10,300,69))\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 122,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "sada\n"
     ]
    }
   ],
   "source": [
    "im.shape\n",
    "print \"sada\""
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 279,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[((610, 462, 128, 19), 'Dmmaeosaundae'),\n",
       " ((608, 434, 201, 22), 'Archaeoptem lithographica'),\n",
       " ((609, 407, 167, 20), 'Conlllclusomls sancms'),\n",
       " ((609, 380, 133, 23), 'Gubmreryx mmum'),\n",
       " ((609, 352, 150, 22), 'Neuquenmm volans'),\n",
       " ((609, 323, 138, 20), 'Concomis Iacustns'),\n",
       " ((609, 295, 145, 23), 'Gamayomis yandica'),\n",
       " ((609, 269, 174, 22), 'Patagaptelyx derenams:'),\n",
       " ((609, 241, 73, 22), 'Bapmrrus'),\n",
       " ((609, 214, 89, 22), 'Hesperomls'),\n",
       " ((607, 184, 148, 22), 'Ausmvis ukhaana'),\n",
       " ((609, 158, 82, 22), 'lcmhyomls'),\n",
       " ((608, 129, 204, 23), 'Aves: Crypturellus undulams'),\n",
       " ((608, 103, 123, 23), 'Aves: Crax pawn'),\n",
       " ((608, 76, 136, 23), 'Aves: Gallus gallus'),\n",
       " ((608, 48, 164, 23), 'Aves: Chauna tomuam'),\n",
       " ((608, 21, 184, 22), 'Aves: Anas plarymyncnns')]"
      ]
     },
     "execution_count": 279,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "labels_for_tree()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": []
  },
  {
   "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": "2.7.10"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 0
}