Cloud OCR SDK Documentation

Arabic OCR SDK by ABBYY

Arabic language has complex script and a lot of peculiarities, which complicate development of Arabic OCR technology:

  • overlapped characters
  • various shapes of characters
  • diacritics
  • variety of Arabic fonts

Here is an example of a text in Arabic:

Arabic text

Arabic OCR in ABBYY Cloud OCR SDK is presented in beta version. ABBYY OCR team constantly collects feedback information from customers and beta-testers and uses this data to improve quality of Arabic OCR. You can use images of Arabic text provided in Sample Images to test Cloud OCR in Arabic for free.

To test Arabic OCR, do the following:

  1. Register to become a user of ABBYY Cloud OCR SDK. If you have already registered, login to the Cloud OCR SDK site and create an Application ID and Application Password. The Application ID and Application Password are used for authentication by an application that uses Cloud OCR SDK. They are passed to the OCR server with each request. See details in Authentication.
  2. Become clear with the type of Arabic documents you are going to process. It can be office documents, books, magazines, or business cards, or even forms. Select the method to be used depending on the type of processing documents:
  3. Set the language parameter of the selected method to Arabic.
  4. Set other parameters of the methods, if necessary. See, for example, How to Recognize Text Fields to know how to tune parameters of a field recognition.
  5. Call the selected method with required parameters and necessary authentication information. The address of the server is http://cloud.ocrsdk.com. The input image file is transmitted in the request body.
  6. Obtain the response as XML. The successful response contains the ID of the created task. For the format of the response see HTTP Status Codes and Response Formats.
  7. Control the status of the task using the getTaskStatus method. When the image is recognized, the response will contain the Completed status of the task.
  8. Download the result of processing. The URL of the result is provided in the XML response.

You can find details on the main processing steps in How to Work with Cloud OCR SDK

Please, also refer to Code Samples. To recognize Arabic text, you only need to modify them by specifying the correct recognition language.