Read handwriting in PDF file

Help
1

I’m struggling with using Read PDF with OCR technology.

Company I work for takes PDF forms which are filled in (by handwriting, inside boxes to be filled) and these are scanned and received to us by e-mail. I need to be able to read these PDF files and the handwriting as best as possible.

How can I do this?

I have tried Read PDF with OCR (Tesseract OCR engine) but the result is looking like:
’ l
. as - I . .

  • i . . ’ I
  • ’“fim flswiaggvereesnt 7 - . Direct Order Form ‘
    ‘ “1.9 ~ Detpanfiemfvetegfisfifiai‘m‘ . RAP Mobility & Functional Support Products I
    77477 77.7.7777717777777777: 7777777777, 7777777777777 777, 777 *7 7(7’7 ’7 77?, 77 1 .7777 ’ “i” 4:7” 77 |
    Provider Hotline Number:
    fioflderygails777 7777 777 7’77 77 7 .7. ’ 7: 777-7 74-
    7 _ 7.0T [Mi 1:] PT 1:] [MO Domer (Specify Profession) ‘7 777 7.7 7’77 7 77 7 7 7‘
    1“ masses,eesgrrebrr‘ - Name LEMLLXEYKE§77 77 7 7 7 . J
    { Providernumber 477777777» 77 7 g .
    ’ r - (RegideredNurseuse fisoeozw 7 7 J .
    .1 - 7 .AHPRAnmnber) 77777177777" 7? 7 i 7 777’- .7 7 77 ——
    g . ’ _ . ’ 'Employer

ETC.

…with lots of 77777’s and none of the actual handwritten components of the form seem to be listed anywhere in the above messy looking string. The PDF file is visibly quite clear to read so I don’t understand why it is coming out so jumbled up.

2

@Jon_G

As you are trying to OCR a Handwritten documents, tesseract OCR not able to parse that

As it require some Machine Learning trained OCR, you could check with Azure Form Recognizer, Abbyy Flexicapture to full fill your requriement

I don’t believe any free OCR will not provide accuracy on handwriting documents

Hope this helps you

Thanks

2 Likes
3

Thanks for your suggestions,

Microsoft Azure (free for max 500 pages per month) might be good enough for what we need to do.

How do I implement using this into my UiPath project workflow exactly?What activity do I use in UiPath to connect to MS Azure Form Recognizer?

I have never done any of this before and fairly new to UiPath itself. I would basically just need to analyse the PDF file and extract certain values such as Names, Invoice Numbers, etc. which are handwritten. Then just feed those string values back into my later activities which get entered into our database.