google-vision-api-notebook.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "name": "Untitled0.ipynb",
      "provenance": []
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    }
  },
  "cells": [
    {
      "cell_type": "code",
      "metadata": {
        "id": "I86tgLoXN-fN",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "import cv2\n",
        "import imutils\n",
        "import json\n",
        "import matplotlib.pyplot as plt\n",
        "import numpy as np\n",
        "import os\n",
        "import pandas as pd\n",
        "import requests\n",
        "import time\n",
        "from base64 import b64encode\n",
        "from IPython.display import Image\n",
        "from pylab import rcParams"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "87ckNcIOiIjf",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "rcParams['figure.figsize'] = 10, 20"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "j1x8BO5AORgS",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "def makeImageData(imgpath):\n",
        "    img_req = None\n",
        "    with open(imgpath, 'rb') as f:\n",
        "        ctxt = b64encode(f.read()).decode()\n",
        "        img_req = {\n",
        "            'image': {\n",
        "                'content': ctxt\n",
        "            },\n",
        "            'features': [{\n",
        "                'type': 'DOCUMENT_TEXT_DETECTION',\n",
        "                'maxResults': 1\n",
        "            }]\n",
        "        }\n",
        "    return json.dumps({\"requests\": img_req}).encode()"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "5PH_sr9TN7eS",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "def requestOCR(url, api_key, imgpath):\n",
        "  imgdata = makeImageData(imgpath)\n",
        "  response = requests.post(ENDPOINT_URL, \n",
        "                           data = imgdata, \n",
        "                           params = {'key': api_key}, \n",
        "                           headers = {'Content-Type': 'application/json'})\n",
        "  return response"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "ia3SZkwhQi8t",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "with open('vision_api.json') as f:\n",
        "    data = json.load(f)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "ziTG4b1-N0oU",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "ENDPOINT_URL = 'https://vision.googleapis.com/v1/images:annotate'\n",
        "api_key = data[\"api_key\"]\n",
        "img_loc = \"Image.jpg\""
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "-RlRp1bllxbh",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "Image(img_loc)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "EU1SEPZWlhGt",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "result = requestOCR(ENDPOINT_URL, api_key, img_loc)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "mfZ_KT68OgV7",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "if result.status_code != 200 or result.json().get('error'):\n",
        "    print (\"Error\")\n",
        "else:\n",
        "    result = result.json()['responses'][0]['textAnnotations']"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "KlGzGqOzROk3",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "result"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "o5AYbd9YfwL-",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "for index in range(len(result)):\n",
        "  print(result[index][\"description\"])"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "iIs5LgaAfwQ6",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "def gen_cord(result):\n",
        "  cord_df = pd.DataFrame(result['boundingPoly']['vertices'])\n",
        "  x_min, y_min = np.min(cord_df[\"x\"]), np.min(cord_df[\"y\"])\n",
        "  x_max, y_max = np.max(cord_df[\"x\"]), np.max(cord_df[\"y\"])\n",
        "  return result[\"description\"], x_max, x_min, y_max, y_min"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Q4LYtrQffm77",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "text, x_max, x_min, y_max, y_min = gen_cord(result[-1])\n",
        "image = cv2.imread(img_loc)\n",
        "cv2.rectangle(image,(x_min,y_min),(x_max,y_max),(0,255, 0),2)\n",
        "plt.imshow(cv2.cvtColor(image, cv2.COLOR_BGR2RGB))\n",
        "print (\"Text Detected = {}\".format(text))"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "f3NoY3tCjDdZ",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        ""
      ],
      "execution_count": null,
      "outputs": []
    }
  ]
}