code/chapter-8/optical-character-recognition/1-get-MSCOCO-validation-image-ids-with-legible-text.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Cloud APIs for Computer Vision: Up and Running in 15 Minutes\n",
    "\n",
    "This code is part of [Chapter 8- Cloud APIs for Computer Vision: Up and Running in 15 Minutes ](https://learning.oreilly.com/library/view/practical-deep-learning/9781492034858/ch08.html).\n",
    "\n",
    "## Get MSCOCO validation image ids with legible text\n",
    "\n",
    "We will develop a dataset of images from the MSCOCO dataset that contain at least a single instance of legible text and are in the validation split."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "In order to do this, we first need to download `cocotext.v2.json` from https://bgshih.github.io/cocotext/."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Archive:  tmp.zip\r\n"
     ]
    }
   ],
   "source": [
    "!wget -nc -q -O tmp.zip https://github.com/bgshih/cocotext/releases/download/dl/cocotext.v2.zip && unzip -n tmp.zip && rm tmp.zip"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Let's verify that the file has been downloaded and that it exists."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "True"
      ]
     },
     "execution_count": 2,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "import os\n",
    "os.path.isfile('./cocotext.v2.json') "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We also need to download the `coco_text.py` file from the COCO-Text repository from http://vision.cornell.edu/se3/coco-text/"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "File ‘coco_text.py’ already there; not retrieving.\r\n",
      "\r\n"
     ]
    }
   ],
   "source": [
    "!wget -nc https://raw.githubusercontent.com/bgshih/coco-text/master/coco_text.py"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "import coco_text"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "loading annotations into memory...\n",
      "0:00:01.462544\n",
      "creating index...\n",
      "index created!\n"
     ]
    }
   ],
   "source": [
    "# Load the COCO text json file\n",
    "ct = coco_text.COCO_Text('./cocotext.v2.json')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "10000\n"
     ]
    }
   ],
   "source": [
    "# Find the total number of images in validation set\n",
    "print(len(ct.val))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Add the paths to the `train2014` directory downloaded from the [MSCOCO website](http://cocodataset.org/#download)."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "True"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "path = <PATH_TO_IMAGES> # Please update with local absolute path to train2014\n",
    "os.path.exists(path) "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Get all images containing at least one instance of legible text"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "image_ids = ct.getImgIds(imgIds=ct.val, catIds=[('legibility', 'legible')])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Find total number of validation images which have legible text"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "3261\n"
     ]
    }
   ],
   "source": [
    "print(len(image_ids))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "In the data we downloaded, make sure all the image IDs exist."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "2752\n"
     ]
    }
   ],
   "source": [
    "def filename_from_image_id(image_id):\n",
    "    return \"COCO_train2014_000000\" + str(image_id) + \".jpg\"\n",
    "\n",
    "\n",
    "final_image_ids = []\n",
    "\n",
    "for each in image_ids:\n",
    "    filename = filename_from_image_id(each)\n",
    "    if os.path.exists(path + filename):\n",
    "        final_image_ids.append(each)\n",
    "\n",
    "print(len(final_image_ids))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Make a folder where all the temporary data files can be stored"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "mkdir: cannot create directory ‘data-may-2020’: File exists\n",
      "mkdir: cannot create directory ‘data-may-2020/legible-images’: File exists\n"
     ]
    }
   ],
   "source": [
    "!mkdir data-may-2020\n",
    "!mkdir data-may-2020/legible-images"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Save a list of the image ids of the validation images"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [],
   "source": [
    "with open('./data-may-2020/val-image-ids-final.csv', 'w') as f:\n",
    "    f.write(\"\\n\".join(str(image_id) for image_id in final_image_ids))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Move these images to a separate folder for future use."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [],
   "source": [
    "from shutil import copy2\n",
    "\n",
    "for each in final_image_ids:\n",
    "    filename = filename_from_image_id(each)\n",
    "    if os.path.exists(path + filename):\n",
    "        copy2(path + filename, './data-may-2020/legible-images/')"
   ]
  }
 ],
 "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": "3.6.9"
  },
  "toc": {
   "base_numbering": 1,
   "nav_menu": {},
   "number_sections": true,
   "sideBar": true,
   "skip_h1_title": false,
   "title_cell": "Table of Contents",
   "title_sidebar": "Contents",
   "toc_cell": false,
   "toc_position": {},
   "toc_section_display": true,
   "toc_window_display": false
  }
 },
 "nbformat": 4,
 "nbformat_minor": 1
}