How to extract data from scanned images using OCR

Hi All,

I am trying to extract data from scanned passport copies but I am unable to use a static procedure to do the same, since based on image the output is getting varied hence I am unable to write a procedure to get data using sub string. Can someone help me here?

Basically I will be having 100’s of scanned copies in a folder and i have to fetch each image get all data and write it into an excel file.

Thanks & Regards,
Delicia.

1 Like

I didn’t try this but hope it may help you

Hi @Delicia_A ,
I am also having the same use case and trying to use Google Vision as suggested by @pathrudu.
I am facing certain issues in adding Google Vision activities package to my studio. Let me know if you plan to try it out and if it helped.

@rinki
did you installed the below dependencies?

  1. Newtonsoft.Json 11.0.2

  2. Microsoft.Rest.ClientRuntime.Azure 3.3.15

  3. Microsoft.Rest.ClientRuntime 2.3.13

  4. Microsoft.Azure.CognitiveServices.Vision.ComputerVision 3.2.0

Regards,
Pathrudu

Hi @pathrudu,

I have tried using the beta computer vision preview version of uipath and it seems to be working fine for working with VMs and for data extraction from online invoices and PDFs.

Is this and the google vision what you are stating is same?

But for passport scanned copies I am not able to extract all the details from it.

@rinki,

You can try using Abby OCR or Computer vision preview version.

Regards,
Delicia.