Using Citrix Relative Scraping U.I path 19.7

Help
1

Hi all

I’m trying to following the exercise here Example of using OCR and Image Automation

using this invoice https://www.uipath.com/hubfs/Documentation/WorkflowExamples/Scanned%20Invoice.pdf

But the UI Path version I have is a later one and I can’t find how to do relative scraping as the menu option doesn’t exist as drop down anymore in the citirx recorder.

Please can someone help me understand how I can scrape the email address using this method so I get the same actions show in the screenshot on UI path page I have given?

Thanks in advance,
Jordan

2

Hello, you can use Recording > Image then this:

Scrape%20relative
Scrape%20relative1133×820 121 KB

1 Like
3

Fine
kindly press alt+ctrl+c being inside the studio
we will be getting popup window where we can find
image

Cheers @jordrowley

1 Like
4

Thanks both and for help you’ve given me in other threads too. I’ll give the whole exercise a go and see how I get on :slight_smile:

1 Like
5

Fantastic
Cheers @jordrowley

6

Hi both, if you get chance, I am struggling to get the total invoice amount of $500. Microsoft OCR won’t recognise any thing and the Google OCR returns all sorts of characters. Please can you see if you are able to get this value if you have the time?

Thanks
Jordan

7

Sure @jordrowley just give me a minute

Meanwhile you could read about new Computer Vision pack 2.0.
It will help you a lot with automating Citrix environments!

8

I am also having trouble with that last amount.

1 Like
9

I’ve found the only way to do it is to zoom out to a size of 71% - it returns the value then using Google OCR. I’m guessing that would mean that if I built a process that always used these invoices, I would need to (or get the robot to) zoom out to 71% before reading them?

Thanks for the info on the new Ui Path options :slight_smile:
Jordan

1 Like
10

Well this invoice was designed to be hard to automate, there is slight chance that you will end up using this kind of invoice in your future.
But your solution to zoom in the PDF is quite intuitive, I will write that down, good job!

1 Like
11

How did you manage to find this google OCR its no more available in 2019.7.0 version

12

It’s called Tesseract OCR now; I assume that it used to be called ‘Google OCR’

1 Like
13

This topic was automatically closed 3 days after the last reply. New replies are no longer allowed.