Azure Ink Recognizer is an AI service in the Vision Category that recognizes digital ink content, such as handwriting, shapes, and ink document layout. Now you can integrate in your applications hand written experiences that are converted to text and made searchable.
Ink recognizer vs OCR
The Ink Recognizer API does not use Optical Character Recognition(OCR). OCR services process the pixel data from images to provide handwriting and text recognition. This is sometimes called offline recognition. Instead, The Ink Recognizer API requires digital ink stroke data that’s captured as the input device is used. Processing digital ink data in this way can produce more accurate recognition results compared to OCR services.
- Handwriting recognition: Recognize handwritten content in 63 core languages and locales.
- Layout recognition: Get structural information about the digital ink content. Break the content into writing regions, paragraphs, lines, words, bulleted lists. Your applications can then use the layout information to build additional features like automatic list formatting, and shape alignment.
- Shape recognition: Recognize the most commonly used geometric shapes when taking notes.
- Combined shapes and text recognition: Recognize which ink strokes belong to shapes or handwritten content, and separately classify them.
Charges are on a per-transaction basis. One transaction corresponds to a request to the service broken down into increments of 200 ink strokes. So, a request of 200 strokes counts as one transaction, a request of 400 strokes counts as two, etc.
The service is currently in public preview and charged at public preview price. GA price will be announced when the service goes GA.
|Free||2,000 transactions free per month|
|S0||$2 per 1,000 transactions|
You can see all the details in the Microsoft Docs and the Ink Recognizer Product page.