Why does Cloud OCR not detecting charecters/numbers from image

  • Last Post 03 February 2014
mpraining posted this 25 January 2014


I am using Cloud OCR SDK in my java app. As you can see from the attached image, the last part of every line is the price, and which is not detected by SDK. With some other images (if necessary I'll upload), such lines are detected as a new line!

I am using the processImage method. I prefer help in terms of java code if possible.

This is the result I got while the following attached image processed. Here you can easily note that the last number in each row is missing in the result. So please help me to find out the issue.

A1020 Roblet-Monnot “Vieilles Vignes� 2010
A1021 Paul Pernot et ses Fils 2008
A1022 Domaine Antonin Guyon 2009
Clos de la Chaume Gaufriot, Beaune
A1023 Domaine Ardhuy 2009
C10-24 Domaine de Lambrays Grand Cru 2009
Clos des Lambrays, Morey
C1025 Camille Giroud Grand Cru 2008
an 18% gratuity is included on all checks


Order By: Standard | Newest | Votes
Anastasia Galimova posted this 03 February 2014

The processImage method has the "profile" parameter. The default profile ("documentConversion") is not suitable for receipt recognition. We recommend to set profile=textExtraction.

If the text order or extra line breaks in the result will not suitable for your task, it could be necessary perform export to XML (set exportFormat=xml). In this format words are saved with its coordinates, and you can process it on your side.

These recommendations on receipt recognition can also be found in Cloud OCR SDK FAQ.

Anastasia Galimova posted this 03 February 2014