I am generating a pdf for a client. This has two columns ... ID and a blank column for the end user to put in some character. I want the client to be able to upload a bmp/jpg/tif whatever (i can specify) image of the printed pdf with its two columns and turn it into a csv/tab/whatever file with the values of the two columns. Can anyone recommend an ocr package that I can access from either a php web PHP or a win32 DLL that I can use from a Windows App (delphi 9)? This doesn't have to be a free package; I am willing to pay. I am aware that you can do forms in PDF and reprocess those ... that is not what I need. I need to OCR a printed document problematically.
+1
A:
Have you seen Tesseract OCR It's C but I assume you can interact in Delphi
renick
2010-06-30 16:08:30
this looks promising! i downloaded and will check it out.
Don Dickinson
2010-06-30 19:36:21
A:
There's a web API for this (pay-per-page, with free trial available) that can generate TXT/CSV from a variety of image formats. It's the WiseTrend OCR API, based on the ABBYY engine. See http://www.wisetrend.com/wisetrend_ocr_cloud.shtml , sign up at http://www.webservius.com/cons/subscribe.aspx?p=wisetrend&s=wiseocr
Eugene Osovetsky
2010-08-11 16:52:40