1

I'm looking for a lightweight software, allowing commercial use, to easily extract alphanumeric fields from images scanned on a flatbed scanner, preferably for Windows (or possibly for Linux).

Not necessarily free, but not costing several hundreds of USD, in the range of 0 to 150 USD.

My need is to extract fields from product stickers (model, serial number, a.s.o.). Texts are black on a white background. The pictures are in PNG format. I can convert them to other image format if necessary. There are no accents in strings. The strings to extract are composed of uppercase letters and numbers. There is no need for dictionaries.

I found the perfect candidate for my needs (IrfanView 32-bit version + KADMOS OCR plugin): http://irfanview.info/plugins/kadmos/, but is only allows personal (non commercial) use.

So, I'm looking for an equivalent allowing professional use.

Requirements :

  • allow professional use, but won't be redistributed
  • be offline, standalone, independant from Microsoft Office and from Google
  • be easy/intuitive to use (not having to call library's funtions)
  • preferably lightweight
  • if possible, allow drag & drop of the image to digitize directly in it, or alternatively allow that PNG files are associated to open in the software
  • allow "drawing" the area where to extract a text string (by drawing a rectangle for instance).

If the software furthermore allows template driven fields extraction, it would be a plus, but it is not required at first.

I'm aware about librairies, but at first, I'm looking for a solution that doesn't require coding.

I tested SimpleOCR which did not convince me.

Other suggestions? Please tell if you tested them or only heard about them.

Thanks.

  • Have you contacted the IfranView dev for pricing? On the 'What is IfranView' page (irfanview.com/main_what_is_engl.htm) he's got an e-mail link where you can request professional use pricing. Given that you know it works for you, and that you are willing to pay some money, you should probably inquire about the pricing - it might not be that expensive. – Michael Kohne Apr 6 '17 at 14:48
  • Michael, the problem is not IrfanView commercial license which I already have and which was cheap. The issue is the license of the Kadmos plugin, which only allows personal use (it displays at install). The products of this editor start at 250 USD: best-ocr.com/prices.html . I'll contact them and ask if they can release a commercial version of the plugin at a lower price. – OuzoPower Apr 6 '17 at 14:56
  • Some suggestions here: techradar.com/news/… Paperport Standard or Paperport Professional are in my price range. I have not tested them yet. – OuzoPower Apr 6 '17 at 15:01
  • Have a look at gImageReader. It is based off of Tesseract OCR. The former is licensed under GPL v3 while the latter is licensed under Apache 2.0. – Mr Public Apr 6 '17 at 16:42
  • Thanks mentioning it. Just tried and the results are very bad. I tried on the sticker of a hard drive. The model "ST500LM000" was read as "sYSunLMnnn", using OCR mode "Plain text" and Page segmentation mode "Sparse text in no particula order". – OuzoPower Apr 7 '17 at 9:37

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.