OCR ROI measures

Discussions about Tesseract OCR integration in GdPicture.
Post Reply
Posts: 1
Joined: Fri Jan 22, 2016 3:10 pm

OCR ROI measures

Post by srbfk » Fri Jan 22, 2016 4:10 pm

We are using GDPicture Tesseract to read scanned invoices. Those invoices have a standardized region, where the invoice code ROI is (bottom right, 2cm from below and from the middle of a A4 page to the right side, defined in millimeters by banks). We tried setting the ROI with exact pixel coordinates, which worked, until some customers started to increase the DPI settings on their printers, which resulted in wrong readings due to wrong ROI coordinates. What is the best practice for such a case? Converting the MM to Inch and use those values as ROI in pixels?

Thanks in advance for any hint!

Posts: 263
Joined: Sun Sep 02, 2012 7:30 pm

Re: OCR ROI measures

Post by Cedric » Mon Feb 08, 2016 10:54 am


Pixels are definitely a bad choice if you do not have the same resolution every time in every document, the most reliable way to work around this is to use a real measurement unit.
The most convenient one is certainly inches because DPI is based on it so if you want to convert to millimeters it can easily become cumbersome (1 dot/inch = 0.03937007874016 dot/millimeter).
Anyway, you can very simply convert your pixels measurements to inches by simply following the well known Pixels ÷ DPI = Inches formula.

Post Reply

Who is online

Users browsing this forum: No registered users and 2 guests