Hello ,as we start to use form recognizer, wanted to know a)what kind of features form recognizer has in terms of pre processing faxed images..does it handle rotated images, portrait vs landscape conversions, removal of grains , faxes taken at some angles. b) Do we have to train for all these variations
@PrasannaCuddalore-7127 Thanks. Please try out prebuilt Invoice, Receipt, and Business Card models and the Layout API using the Form Recognizer Sample Tool. See how your data will be extracted without writing any code.
Please follow the doc for recommended input requirements.
If you are extracting only text, tables and selection marks from documents you should use layout, if you also need to extract key value pairs you can train a custom model or use a pre-built (Invoice, Receipts, Business Cards). Layout results (text, tables and selection marks) are included in all the Analyze outputs (custom and pre-built) in the readResults (text) and pageResults (tables) of the JSON output.
• Layout – extract text, tables selection marks no training required
• Pre-built – Invoice, Receipts, Business Cards – extract values of interest from these type of documents
• Custom – Extract key value pairs trained on your own documents
All of the above will also include the text, tables and selection marks in the results.
thanks for getting back, we are using supervised learning by labelling data using some sample forms.. We are satisfieying all requirements in your link, but had questions around the below..
however in reality some of the data we get have faxed images with either
a) gray scale
b) rotated tiffs
c) sent at an angle
does form recognizer do any kind of pre processing out of the box in that it removes gray scale, enhances the image removing dots, gray spots etc that are usually seen in faxed data ?
pls advise
thanks
@annapras-7127 Thanks for the details. We are checking with the product team to confirm on the pre-processing of the same.
3 people are following this question.
