I’m using a Find OCR Text Position on a website to look in a vertical list of the alphabet and find each letter. I’m having it highlight the text location to confirm it’s finding the correct spot. It has found every letter successfully except “O”. When looking for O It finds nothing and nothing gets highlighted. I was able to use the “get text” activity on the “O” and confirmed it is an O.
Find OCR Text Position takes a screenshot and uses OCR to identify the text, so it only sees the image and not the actual character.
Since “O” looks very similar to “0” (zero) or other round shapes, OCR might not detect it correctly. Try adjusting the OCR settings to improve accuracy.
Could you also share a sample image? It will help us understand the issue better.
Hi @charlotte.e,
Since Get Text returns O but Find OCR Text Position cannot locate it, the OCR engine may be interpreting the character differently for example, as 0 instead of O.
Try using Get OCR Text on that area to verify what the OCR engine is reading. You can also try a different OCR engine (Microsoft OCR, OmniPage, Tesseract, etc.) or adjust the OCR scale/accuracy settings. If this is a web element, using selectors would be more reliable than OCR.
Thanks
Hi @charlotte.e
This is likely an OCR recognition issue. Try changing the OCR engine, increasing the scale factor, and adjusting accuracy settings.
Happy Automation
if you are able to get text(natively) then why to use OCR?
use get attribute(“text”) of the HTML element and then compare..
also why do you need text position?
could you add some screenshots and elaborate on what you want to achieve so that we can see if non-ocr actions could help you.
most OCRs have issue in recognizing characters like “0/O”, “i/I/1” etc.. so i would take OCR as my last resort.
SG.
Try increasing the browser zoom level or screen resolution and test again. Also you give a try with different OCR engine (e.g., UiPath Screen OCR, Microsoft OCR, OmniPage OCR) and compare the results.