Document Recognition with email Attached Images

#1

It seams to me as a common problem, but I just could not find any Activity that address this question. I found them for reading emails and saving attachments, but thats pretty easy.

The main part, that is my greater interest, is how can I recognize that the attached images (jpg, png, …) are of a specific kind of information (lets say, a Drivers Licence) and then extract the name and id of the driver.

What kind of Activity can I use for that ?

1 Like
#2

Hi,
As you know how to extract from email and save the attachments in your local and get the file name.
Once received the files you can use various ocr engines to get the required fields that you want or you can use ABBYY flexicapture to get the fields, create doc definition and then store to excel or database and fetch from there using various activities in uipath.

Regards,
Pavan H

2 Likes
#3

But how do I recognize what kind of document I am receiving? Making it more clear, supose I receive two attached files like these you can see in theses links (https://radionajua.com.br/gthumb.php?h=9000&w=710&src=media/noticias/paulosava/14640923520.89379500.jpg, https://static.poder360.com.br/2018/02/dni.jpeg) - two kinds of documents used in Brasil (Citizen ID and Worker Registry).

I supposed there were something that would recognize some graphics elements (a logo, for example), but I did not find it in OCR activities.

#4

Hi,
You can use if condition insid for each loop of output of save attachment activity to check what type of file it is. Or if you want to use abbyy flexicapture, you need to know the different types of files you will be expecting and for each of those sample files you need to create a template to extract fields and in project set up station create project for all the types of templates created.

Let us know if this helps.

Regards,
Pavan H

2 Likes
#5

Where do I find Flexicapture ? It is not present in OCR activities.
I found in another post that we have to install Intelligent OCR Activities.

I tryied that, but the package does not install.

Anyway, reading the documentation, I still can´t find solution for what I need. It is not only the type of file (PDF, JPG, etc). It is image recognition.

#6

Forgot to mention the problem:

Failed to download package ‘Numsense.0.12.0’ from ‘https://www.myget.org/F/workflow/api/v2/package/Numsense/0.12.0’.
The HTTP request to ‘GET https://www.myget.org/F/workflow/api/v2/package/Numsense/0.12.0’ has timed out after 100000ms.

#7

Hey,
Not sure what is your requirement but what I usually do is: I am using flexicapture studio which is a paid version and then from there I am sending the data to a database and using uipath I am using database activities to get the data into uipath

Let us know if this helps you.

Regards,
Pavan H

#8

Thanks for your attention. Understood your approach. I am doing a prof of concept, so not the moment to buy licences. Looking for RPA as a auxiliar tool for BPM process in Bizagi. Thought UIPath had image recognition features, but seems that it requires other tools.

I need something like this:

Read mails + attachments -> identify diferent types of documents (attached pdfs ou images), classify them and get OCR information from them -> Create case in Bizagi with information classified.