tutorial-recognize-text.md

copyright	lastupdated	keywords	subcollection
years 2015, 2019	2019-06-04	Text recognition,Visual Recognition beta Text model,Text model,recognize text	visual-recognition

{:shortdesc: .shortdesc} {:external: target="_blank" .external} {:tip: .tip} {:important: .important} {:pre: .pre} {:codeblock: .codeblock} {:screen: .screen} {:apikey: data-credential-placeholder='apikey'} {:url: data-credential-placeholder='url'}

Recognizing text in an image

{: #tutorial-recognize-text}

This tutorial guides you through how to make your first call with the {{site.data.keyword.visualrecognitionshort}} beta Text model to detect and recognize English text in an image. {: shortdesc}

The Text model is a private beta feature, and you must have permission from {{site.data.keyword.IBM_notm}} to make calls to the model. Request access{: external}. For more information about beta features, see Release notes. {: important}

Before you begin

{: #tutorial-recognize-text-prerequisites}

1. Go to the {{site.data.keyword.visualrecognitionshort}}{: external} page in the {{site.data.keyword.Bluemix_notm}} Catalog.
   1. Sign up for a free {{site.data.keyword.Bluemix_notm}} account or log in.
   2. Click Create.

• Copy the credentials to authenticate to your service instance:
  1. Click Show to view your credentials.
  2. Copy the API key value.

Step 1: Recognize text in an image

{: #tutorial-recognize-text-recognize-text}

1. Issue the following call to recognize text in an image{: external}. Replace {your_api_key} with the API key value that you copied earlier.

   curl -u "apikey:{your_api_key}"{: apikey} \
   "https://gateway.watsonplatform.net/visual-recognition/api/v3/recognize_text?url=https://watson-developer-cloud.github.io/doc-tutorial-downloads/visual-recognition/lookButDontTouch.jpg&version=2018-03-19"

   {: pre}

   The response includes the recognized text and the locations of words in the text with a confidence score for each word. The location information can be used to draw bounding boxes around the words.

   {
     "images": [
       {
         "image": "lookButDontTouch.jpg",
         "text": "look but\ndont\ntouch",
         "words": [
           {
             "word": "look",
             "location": {
               "height": 651,
               "width": 1235,
               "left": 914,
               "top": 1591
             },
             "score": 0.9718,
             "line_number": 0
           },
           {
             "word": "but",
             "location": {
               "height": 651,
               "width": 939,
               "left": 2148,
               "top": 1591
             },
             "score": 0.9246,
             "line_number": 0
           },
           {
             "word": "dont",
             "location": {
               "height": 586,
               "width": 1594,
               "left": 1220,
               "top": 2240
             },
             "score": 0.5823,
             "line_number": 1
           },
           {
             "word": "touch",
             "location": {
               "height": 629,
               "width": 1701,
               "left": 1193,
               "top": 2824
             },
             "score": 0.8806,
             "line_number": 2
           }
         ]
       }
     ],
     "images_processed": 1
   }

   {: codeblock}

Next steps

{: #tutorial-recognize-text-next-steps}

You have a basic understanding of how to recognize text in an image. Explore further.

• Read the overview.
• Explore the Text model methods in the API reference{: external}.

Attributions

{: #tutorial-recognize-text-attributions}

• lookButDontTouch{: external} by Lubo Minar{: external} on Unsplash{: external}. No changes were made to this image.