Functioning of automatic receipt recognition (OCR)

The feature of automatic receipt recognition allows to save time when creating an expense report. 

The text on the receipts (pictures of receipts, files) is recognized to pre-fill in the expense. The user will have to check the expense and complete it. 

Then, the OCR (Optical Character Recognition), allows to translate a picture in words.  

 

Who is concerned by this article?

All users of Eurécia. The feature is, by default, active for all users of the Expense Reports Module. 

 

How works the OCR?

Enable or disable the OCR

To Enable or disable the feature:

• Click on  
• Check or un-check 'Supporting document automatic recognition' 
• Then, this choice will be saved for the next expense reports 

To disable the feature:

• Create a new expense
• And click on 'Disable' :

 

Use the OCR

• Create an expense report or open an existing one 
• Create a new expense
• Add a receipt to the expense 
• Some fields are automatically identified and are filled-in automatically, in green: 

• Check and complete the information requested 
• Save the expense

 

Informations recognized by the OCR

 

Fields that can be pre-filled in are: 

• Date
• Expense type 
• Total (tax-inclusive)
• VAT
• Currency

 

 

How works the OCR ?

The Eurécia OCR uses keywords present on the supporting documents to determine the type of information to be recovered. For example, it will detect the terms 'Restaurant', 'Lunch', 'Toll', etc. 

These keywords are also used to find the type of expense among those set up in the expense profile.

Once the expense type has been found that the processing of the other information begins: the date, the amount including VAT, VAT and the currency.

Caution !  If the identification of the expense type has not worked, Eurécia fill-in the expense with the date of the day and you need to fill-in the information manually.

 

FAQ

Will the addition of the receipt after the expense has been completed overwrite the manual input?

No.

If the user has already completed the expense, adding a receipt will launch the identification. But, only the empty fields will be automatically filled-in.

The OCR won't work if the user has already completed every field that can be identified by the OCR.  

 

Can I manually relaunch the OCR?

The automatic recognition cannot be manually relaunch. 

Only the addition of a receipt to a new expense activates it.

However, it is possible to add the receipt at any time during the input. 

 

Does the OCR launch itself for global receipts of the expense report? 

No, the OCR launches only for the receipts in each expense. 

 

What are the recognized formats?

The files formats that will be displayed on the screen and that can be recognized by the OCR: 

• JPG / JPEG
• PNG
• GIF / Animated GIF
• BMP

 

PDF files will be displayed on the screen but won't be recognized by the OCR. 

The optimal size for a picture is 1024*768 and the maximum size is 4 Mo.

 

Recognized date formats: : 

• 31-12-2019
• 31/12/2019
• 31.12.2019
• 31/12/19
• 31-12-19
• 31.12.19
• 31JAN20
• 31 January 2020
   

Keywords associated to this article:

Expense reports - Reimbursement - professional expenses - receipts - OCR - Optical Character Recognition