Get ocr text issue

[A_Learner]

Hi,
I am using Microsoft OCR in my get ocr text. It is not reading text accurately. Some letters like S and T are left out. Other OCR methods are asking for API key. Trseract is giving me an error. Can you suggest a bette and free OCR please?

Thanks

[Parvathy (PS Parvathy)]

Hi @A_Learner

Try Omni Page OCR. you need to download the dependency UiPath.OmniPage.Activities.

Go to Manage Packages → All Packages → Type the package name → Click Install → Click Save.

Hope it helps

[A_Learner]

Thank you so much! Will certainly try this.

[mkankatala (Mahesh Kankatala)]

Hi @A_Learner

Extraction is all based on the OCR selection, Each OCR will extract differently and give the output text in different patterns. Try the different OCR’s we have in UiPath Studio, for the Structured documents use the UiPath document OCR, for the unstructured or semi structured documents use the Teserract or Omnipage OCR, for using Omnipage you have to install the UiPath.OmniPage.activities package.

If you are getting the same problem, for every OCR there is a property called Scale as shown in the below image. First give the scaling value as 2 change it to 2.5 and increase the values.

Hope it helps!!

[A_Learner]

Thankyou @mkankatala and @Parvathy
Increasing scale has solved my issue. How much can I increase it to? Right now, I set it to 3.

Thanks,

[mkankatala (Mahesh Kankatala)]

Thank you @A_Learner

You can increase up to 5, but don’t rush to increase by point, change it by .5 up to 5.