‚Äč

Cloud OCR SDK Documentation

Frequently Asked Questions

Why cannot I login to my account?

Please try again (https://cloud.ocrsdk.com/Account/Login), checking for possible mistakes:

  • Make sure that you are using the correct e-mail and password.
  • One possible reason is that some extra symbols (like spaces or line breaks) were copied with the password. To avoid this, copy the password to a text editor and from there to the form.
  • Double-check the symbols that look alike (for example, the capital letter I and the lowercase letter l).
  • Make sure that Caps Lock is off.

If you are still unable to log in, try resetting the password by clicking Forgot password? on the login screen.

Where do I find my Application ID and password?

Your Application ID is the name which you entered when creating the application. You can see it if you go to http://cloud.ocrsdk.com/ and log in:
Application ID on your dashboard

The password to your application will be sent to the e-mail address you specified during registration. See also the next question.

I didn't receive an e-mail with the password for my new application. What should I do?

First of all, check that the e-mail didn't end up in your spam folder. If not, try resetting your application password to receive another e-mail with a new password. To do this, go to http://cloud.ocrsdk.com/ and log in. Click Settings under your application's name. On the next screen click Reset application password.

Why do I receive e-mails from Cloud OCR SDK?

The e-mail notifications that you receive from the Cloud OCR SDK team contain information that can be useful for you, such as the new password for your application or a reminder that your application's balance is running low. See Service Notifications for details.

How do I convert a personal Cloud OCR SDK account into a corporate one?

If you need several members of your team to manage the account that was previously a personal one, we recommend creating a separate group email which every of these persons will be able to access. Then please send an email request to CloudOcrSdk@abbyy.com and we will change the account email. After our database is updated accordingly, a message with a new password will be sent to the new email.

How many applications can be created for one user?

You can create as many applications as you want, as long as you have purchased at least one package for each of them, excepting the first one.

How do I purchase a recognition package for my application?

You can buy a package online. Payment is performed via PayPal, either with your credit card or your PayPal account. See Online Purchase Guide for details.

How can I get PDF invoices for my purchases?

To receive PDF invoices for your purchases, please enable the option "Send PDF invoices for every completed purchase" and fill in the invoice settings in your Account Settings. In this case the invoice will be attached to the confirmation e-mail of every new purchase.

You can get the invoices for the previous purchases in the Order History section. Click on the link in the Invoice# column.

I recognize documents in different languages and don't know exactly the language of each document. What should I do?

You can specify several languages, separated by commas, in the language parameter of any processing method. For example: "English,Spanish". The recognition service will automatically select the language of the document from the specified set.

Can I receive the recognition result in several file formats?

Yes, you can export the recognized data in up to three different file formats. To do this, specify the necessary formats, separated by commas, in the exportFormat parameter of the processImage or processDocument method. For example: "txt,pdfSearchable,xml". The task cost does not go up when you set multiple export formats.
Note that you cannot export in multiple formats when using a field-recognition method or business card recognition method.

How can I use Cloud OCR SDK service to recognize receipts?

You can recognize the receipt image and receive the recognition result in XML format, which contains not only recognized text but also the text blocks' coordinates. Then extract information from this XML on your side, using the coordinates to restore the receipt structure. See Output XML Document for details.
We recommend the following settings for receipt recognition:

  • disable any geometrical correction to keep initial word coordinates: set correctSkew=false, imageSource=scanner,
  • set textExtraction profile to enable detection of small text areas.


However, specialized receipt recognition functionality is also planned for Cloud OCR SDK in future.

Why the output of field-level recognition does not contain all the text of the field?

The technical restriction on the text field length is 200 characters. If the field you recognize is longer, the extra characters will not be written into the output. If you need to process a lot of text, it is better to use the processImage method, which provides full-page recognition.

Why cannot my PDF file be processed? The “File … has access restrictions and cannot be added to the document” error is returned.

This error message is returned when you try to process a PDF file which is protected by password. If you know the password, specify it in the pdfPassword parameter of the processing method. If you do not know the password, try a workaround: if printing is allowed for the file, print it with some virtual printer software into one of the image formats accepted by ABBYY Cloud OCR SDK and process images instead of PDFs.

My images are processed too slow, what's wrong?

The processing time depends on many factors, the most important being your image quality, the upload speed, and the service workload. Here are some things you can try if you are dissatisfied with the processing speed you get.

  1. See Source Image Recommendations for advice on scanning and photographing your documents so as to improve recognition quality and save processing time.
  2. Check your upload speed, especially if you are sending large images.
  3. If your images are of good quality, you can skip some of the preprocessing stages to increase recognition speed. Set up the following parameters of the processing method you are using:
    • imageSource=scanner
      if your images are scanned and not photographed; some types of distortion will not be detected
    • correctOrientation=false, correctSkew=false
      if your image is the right side up and not askew
  4. If you do not need barcode recognition also set readBarcodes=false.

See Service Scalability for details about how the workload is distributed and the service scaled up when the processing queue gets too long.

How can I check if the service is operating normally?

Go to http://status.ocrsdk.com/ to check the current service status and find out more about the past performance issues.

Why do I get 'Error 450: Exceeded quota to add images' while testing Cloud OCR SDK performance?

To prevent DoS attacks ABBYY reserves the right to impose reasonable limits on the number of free repeated recognitions per account per day.

If you get the 'Exceeded quota to add images' error, your application has run up against one of these limitations:

  • you may upload twice as many images as your current balance would allow you to process (you can resolve this issue by topping up your account);
  • up to 100 images and up to 100 barcodes per day can be re-recognized for free (if you have already re-recognized more, please wait until the next day).

See Billing Terms for details.