web
You’re offline. This is a read only version of the page.
close
Skip to main content

Announcements

News and Announcements icon
Community site session details

Community site session details

Session Id :
Power Platform Community / Forums / Power Automate / OCR issues for some co...
Power Automate
Answered

OCR issues for some comboboxes

(0) ShareShare
ReportReport
Posted on by 23

Hello there,

 

Does anybody else having problems with some comboboxes while using the OCR capabilities in Power Automate Desktop? I don't know why but it worked for another combobox just before failing with this other one and i don't get it honestly. I'm using the "move cursor to text" feature, and I'm doing everything that i found on the forum before writing this message, which includes increasing image to 2-3 for the ocr to work better, limit the 

 

DavidValencia_0-1659701197119.png

 

I have the same question (0)
  • AS-06081019-0 Profile Picture
    1,738 Super User 2024 Season 1 on at

    It could be the background blue color causing the issue, try taking an image without selecting 2022.

     

  • DavidValencia Profile Picture
    23 on at

    Hello @Ankesh_49,

     

    The problem is that, once i've pressed the button that makes this window pop up, it already shows up highlighted or selected. It happened as well for the other combo.

     

    DavidValencia_0-1659942773752.png

     

  • Pavel_NaNoi Profile Picture
    1,074 on at

    OCR is tricky as hell, I'd recommend to use UI Elements extraction methods instead if possible, I've had issues where OCR can easily find any time text from 00:00 - 59:99 from a perfect image, but can't find 1:00:00 - 99:99:99 for the life of it.

     

    I'd say in your case you probably need to have different OCR image widths & heights for different combo boxes as they have different font sizes.

  • Verified answer
    DavidValencia Profile Picture
    23 on at

    I captured the exact pixels where the information begun, but while debugging with "Extract text with OCR" i found that, even if you captured the search area precisely, said method would take more circundant information, that is 10 pixels more to the right and 10 less to the left (X-axis). The Y-axis was alright, nontheless i gave it a little more room to work and it helped a lot performance wise, no need to read the entire screen again, and precision wise, even if it detected the searched text in the entire screen sometimes it would confuse words with ones in the background and so.  

Under review

Thank you for your reply! To ensure a great experience for everyone, your content is awaiting approval by our Community Managers. Please check back later.

Helpful resources

Quick Links

Season of Sharing Community Challenge Winners!

Congratulations to our community stars!

Kudos to our 2025 Community Spotlight Honorees

Expanding mentorship, skilling, and AI innovation

Congratulations to the June Top 10 Community Leaders!

These are the community rock stars!

Leaderboard > Power Automate

#1
David_MA Profile Picture

David_MA 252 Super User 2026 Season 1

#2
11manish Profile Picture

11manish 169

#3
Haque Profile Picture

Haque 154

Last 30 days Overall leaderboard