How can we improve OCR output accuracy quality?

Hi, I'm using ABBYY's Mobile SDK for detecting text on an image.

I'm using passing image bitmap to 'IRecognitionCoreAPI'. How can I improve the accuracy? and avoid junk characters.

Thank you.

Was this article helpful?

0 out of 0 found this helpful


1 comment

  • Avatar
    Aleksandr Egorov

    Hi, the overall accuracy depends highly on the image you're passing to CoreAPI.

    • If you're using video stream to take photos then we recommend using IDataCaptureService
    • For CoreAPI you can tune only TextTypes. If it doesn't help then I would ask you to create a request sending the image[s] for investigation to us.

Please sign in to leave a comment.