Azure Cognitive Services OCR

Azure Cognitive Services OCR

Well, here we will discuss, a very important topic i.e. Azure Cognitive Services OCR. At the end of this article, you will definitely get some idea about this particular topic.

Azure Cognitive Services OCR

OCR means Optical Character Recognition (OCR) which actually helps you to detect the different text in an image and then extracts the different recognized characters into a machine-usable format. The Computer Vision API of Azure includes Optical Character Recognition (OCR) capabilities.

In real-time scenarios, when you have the requirement that users will just upload an image and you need to read all the text values from the image, and then you need to fill a form from the different values that you got it from the image. In this type of scenario, the Azure Cognitive Services OCR really helps the developer a lot, We all know that this is not an easy job but with the help of the API, it is really easy to achieve this type of functionality.

While working with the OCR, in the successful scenario, you will get the output as expected but in case of failure, you might get the messages below

  • InvalidImageFormat (Input data is not a valid image.)
  • InvalidImageSize (Input image is too large.)
  • NotSupportedImage
  • NotSupportedLanguage (Specified language is not supported.)
  • InvalidImageUrl (Image URL is badly formatted or not accessible.)

Or, You might get the error code 500 and the issue can be as below

  • InternalServerError
  • FailedToProcess (Failed to process the image.)
  • Timeout (Image processing time out.)

Remember, in the case of the ACS OCR, Below are the input requirements, which are a must.

  • The file format must be JPEG, PNG, BMP, PDF, and TIFF
  • It supports up to 2000 pages in the case of the PDF and TIFF files being processed. One more thing is for the free tier, only the first two pages will be processed.
  • Remember that it is 50 MB and in the case of the free tier, it is 4 MB. The dimension must be within 50 x 50 pixels to 10000 x 10000 pixels.
  • At max 17 x 17 inches dimensions must be for the PDF and the size will be like A3 paper size or smaller.

The API supports reading the text from the image, text from a document, Text from handwritten documents (Only English language is supported as of now), Text from printed documents, and Text from Mixed language documents.

These are a few key points on this topic that we should know while working on the implementation of it.

As part of the implementation of Azure Cognitive Services OCR, You can check out the example How To Extract Image Details Using Azure Cognitive Service. Where we were able to read the text from a computer-generated bill.


In this article, we discussed Azure Cognitive Services OCR. Thanks for reading this article !!!