Fixed Form Data Extraction selecting Text over Date in Date Field

Help Activities
, ,
1

Hi All,

I’ve got a curious case where I have a clearly labelled date field which is consistently selecting a Text value over a date value

This is the settings I’ve applied in the Taxonomy

Hub24_UipathForum_Image1
Hub24_UipathForum_Image11132×401 24.5 KB

Here is the selected values in the Document
image

And the detected options
image

Is there a simple way to force the Data Extraction to select the actual date data over the text data?

For additional information, the Extract model is an Old Fixed Form extraction model, which appears to have lost its “Template” document a while ago which means I can no longer update it. I’ve added in a couple of additional Fixed Form extraction models, however these don’t appear to be getting selected over the old extraction model

2

@Stefan.He_Enlift,

Try changing OCR Engine and Apply OCR on PDF.

image
image687×774 16.4 KB

image
image835×388 34.5 KB

Thanks,
Ashok :slight_smile:

3

I gave that a go, but unfortunately its made my results even worse

image

image

4

@Stefan.He_Enlift,

Which OCR engine you are using?

5

My original settings were

  • UiPath OCR Engine (For Digitization and all Classification/Extraction training)
  • ApplyOCROnPDF = Yes (For Digitization, and all Classification/Extraction training)

I tried out the Tesseract OCR engine which is free to use specifically for a new template. And also tried variations of “ApplyOCROnPDF” = Yes and “ApplyOCROnPDF” = Auto with no luck.

6

@Stefan.He_Enlift

looks like your indicated region also looks wrong…can you try checking that

cheers