OCR_Text Recognization

Help
1

Hello All,
I am trying to capture text using OCR, but the text is dynamic and changes every day.
How can I make it stable to capture and extract the text whatever it is.

PS: The OCR is not able to give 100% result, Such as

Instead of 1, it is resulting in I,
Instead of 0, it is resulting in o,

I tried by replacing that with the desired one, but as the text is not fixed, so can not make it stable.

The line which I want to capture contains:

This folder contains 52 documents: 1 from XYZ/ABC and 51 from JKLM.

Here the bold text represents the dynamic text which changes for every process.

Please find the below image for reference.

Skype-20190719-215609
Skype-20190719-2156091774×1004 144 KB

Thanks in advance. :slight_smile:

2

@kgmaheshwari Have you tried increasing the Scaling property of read OCR text activity.

3

@WhenCutEsh Yes Venkatesh, I’ve tried that, but that works for handling text from 1 to I, 0 to o only, and that too not every time. along with that, I am looking for some solution to make it stable for all the dynamic text as well.

4

@kgmaheshwari If you can provide few sample images We can try using different OCR’s.

5

for an example to capture the random 1st line from this webpage;

after pressing go, the text changes

6

@kgmaheshwari With Microsoft OCR I 'm able to achieve your requirement. Try it.

7

Can you please share the sample workflow