Hello All,
While using 'Read text with OCR' in Surface Automation, the data is getting misinterpreted. We have tried all the scaling options from 1 to 14 and all the options available in the Read text with OCR but nothing is giving 100% accuracy.
Every time, we will be receiving the data dynamically when we read from windows based application.
For example,
When we are reading the data from windows application, we are getting the result data as JSSN in the place of SSN , O or o in the place of 0, L in the place of 1, ? in the place of 2, getting the spaces in between the text and some times we are getting the empty values as well.
Please refer the attached screenshots for result data using Read Text with OCR Action.
------------------------------
kavya shrie
------------------------------
HI Kavya Shrie,
Not sure it is only for me but I'm unable to see the screenshots.
Is there a reason for why you opted Read text with OCR ?
Did you tried Recognize text?
check the Surface automation guide
------------------------------
-----------------------
If I answered your query. Please mark it as the "Best Answer"
Harish M
Lead developer
America/New_York TX
------------------------------
Hi
Thanks for the response.
We are trying to extract the data from Windows based(Smartdesk) application, through normal spy modes, we are unable to extract the conventional mode. So we are trying with the surface automation. These are the details we are trying to extract from the smartdesk application:
In the surface automation, we have spied the elements and recognized the font as "Tahoma 10" and through read action with the Recognize Text Property, we are getting the below results :
Next, we have tried with Read text with OCR(Tried with different scale ratios as well) and we are getting the results as below with some special characters and with mismatches:
Please help to give suggestions to better the results. Thanks in advance.
------------------------------
kavya shrie
------------------------------
