Sample Workflow using document understanding and uiPath document OCR engine

[MariaJosephina]

Hi,

Can anywhere share few sample Workflow that uses ‘uiPath document OCR’ engine to digitize the data and extract certain data from the scanned arabic and english documents.

[supermanPunch (Arpan)]

Hi @MariaJosephina ,

The Template of Document Understanding Process should be available in Studio Itself.

It uses the UiPath Document OCR Engine and has some document examples.

However, we would need to configure it with all the requirements before it can be used.

I am not sure about arabic language detection but it does extract data in english.

You could give it a Try and let us know what is the Main Outcome required for your case.

[MariaJosephina]

Thanks for the reply.

Template is not found in my studio.

[supermanPunch (Arpan)]

@MariaJosephina ,

Maybe you do not have the Latest Version of Studio.

You could also find it UiPath Marketplace.

But I will just upload the Template that I have created from Studio :
DocumentUnderstandingProcess.zip (3.7 MB)

[MariaJosephina]

Thanks a lot for sharing the workflow.
DO we require separate license for uipath document OCR engine and for other data extraction methods (like Machine learning, intelligent form extractor)

Also is it possible to specify the lanaguage for uipath document OCR engine.
(For example if we want to extract the arabic text, is there a provision for specifying the language as “Ara”)

[supermanPunch (Arpan)]

@MariaJosephina

DO we require separate license for UiPath document OCR engine and for other data extraction methods (like Machine learning, intelligent form extractor)

@MariaJosephina , Yes, We would require the Document Understanding License.

For Machine Learning, I suppose there is a Separate License.

However, In Automation Cloud you can Request for a Enterprise Trial and make the features available to you.

@MariaJosephina

Also is it possible to specify the lanaguage for UiPath document OCR engine.
(For example if we want to extract the arabic text, is there a provision for specifying the language as “Ara”)

I am afraid, there is no such property to specify the language for extraction with this OCR Engine.

However there are other engines which you could try extracting from Like OmniPage, Microsoft and observe the extraction Results.

[MariaJosephina]

Noted.

For machine learning extractor, are we using the same api key for document understanding (enterprise trial)

[supermanPunch (Arpan)]

@MariaJosephina , We use the Same API Key for the Machine Learning Extractor as well.

However, the endpoints for each type of the Machine Learning Extractor/Machine learning package may be different.

Below Doc contains the information about the endpoints for ML Packages :

Public Endpoints

[MariaJosephina]

Noted.

Thanks for the support.
I would also want to understand, if I want to add Abby OCR engine to Get OCR text activity/ Digitize activity of Doc_Understanding, what steps must be done to proceed with this. (Should I download Abby/ should we use api key)