There are multiple ways you can achieve it, you would need to define these as templates and execute with the help of an document extraction engine… it can not be done effectively with plain vanilla OCR… here are few pointers to get started…
1- UiPath Document Understanding (Receipt and Invoice AI - Now available in Public Preview!)
2- UIPath with Rossum AI solution (Rossum and UiPath: Invoice extraction in 15 minutes | Rossum)
3- UiPath with Abbyy Flexcapture (Making Robots Smarter - ABBYY and UiPath)
I hope its helpful for your pursuit… I have implemented all 3 approaches with sample templates… will be happy to answer any questions you may have…
I have made sample solution for you with UiPath Document understanding model.
your invoices are in the input folder and it extracts invoice number , date and total to an excel.
If you want you can add more fields to extract. I have also commented out the validation station you can uncomment and check if you need. Also test it with more invoices files inside the input folder. Hope it helps you to start off. InvoiceReading.zip (44.5 KB)
Thank you for the sample!!
I’ll try it.
I tried your solution and its flawlessly.
But when I tried the document understanding other day ACCURACY was very low.
Is there a particular reason for this?
Which OCR technology you used in Digitize section?
Can you please share the code which you tried?
I can have a look.