Session expiration Your session is going to expireClick here to extend

Budget (€) :

Small project <800

Posted on

1/24/18 1:22 PM

This project has expired

Why don't you register anyway? We are sure that you will find many similar projects out of the thousands waiting for you!

Post similar project now


Operating system: Android and iOS

Prerequisite: Integration into an existing app must be possible


1.) Different invoice types (max. 20) are to be read out via photo from the content and interpreted correctly.

2.) Each individual item position is assigned to the best search match so that a price comparison is automatically made possible (between point 1 and the search result)


There is already an app (name on request), which we would like to extend with a function. The goal must therefore be to make such an integration possible via interfaces, libraries,.... in the end. The conversion has to be done for Android and iOS. The apps are about a search that returns items for purchase.

Now the customer should be able to take a picture of an invoice (possibly several pages). These are to be recognized and interpreted via OCR. An integration of external OCR recognition software (e. g. Google) would be conceivable. The invoice contains the article number, description, quantity, content, packaging, unit price, total price (net). This information is needed for the search and must be interpreted correctly. The different invoice types can be identified by the respective company logo. We are talking about a maximum of 20 different templates, where the position of the columns can vary. With the correct interpretation (point 1), all data are then available for the search.
The storage of this data for a future faster article allocation (point 2) can be helpful. Use the Total price/quantity fields (and if quantity is empty by content) to determine the unit price. Depending on the template, the article description must be broken down differently and evaluated with a data basis to get the best result in that template-case. The result of this research is an article number which is possible to search for in a shop (for the fields price and article description).
The customer is then to be informed of the selected and found articles for each item on top of each other:
Article ID acc. to RG | Designation acc. to RG | Quantity acc. to RG | Unit price acc. to RG | Total price acc. to RG
Article ID acc. to agreement | Designation acc. to agreement | ___ | Unit price acc. to agreement | Total price acc. to agreement

The total price is then displayed under all items:
Total price according to RG
Our total price (addition of the total article prices)

To descripe it in short words:
Customer has to Scan and with that the result ist he pricecomparison, so that he knows what he would have to pay there.