I believe tesseract-ocr is also based on the ocr work out of HP labs, same as gocr. I would take those numbers with a lump of salt.
That will probably be true for text in a single column, no font size changes or type face changes. On 6/11/08, eric pareja <[EMAIL PROTECTED]> wrote: > how does tesseract-ocr fare? [http://code.google.com/tesseract-ocr] > > character accuracy is about 98%, word accuracy is 95+%. > > On Tue, Jun 10, 2008 at 10:35 PM, Paolo Falcone <[EMAIL PROTECTED]> > wrote: >> There's no "poor man's OCR". The current state of GOCR (jocr.sf.net) >> is just so pitiful at this stage, it's not worth even considering. > > -- > `..^..' eric pareja ([EMAIL PROTECTED]) lpic-2 | software freedom for > all > |<(e)>| gnu linux python debian edu iosn localization tagalog filipino > `..v..' foss internationalization usability pusakat philippines free > "Ang mundo ay aklat, at iisang pahina lamang ang nababasa ng hindi > naglalakbay." > _________________________________________________ > Philippine Linux Users' Group (PLUG) Mailing List > http://lists.linux.org.ph/mailman/listinfo/plug > Searchable Archives: http://archives.free.net.ph > -- Sent from Gmail for mobile | mobile.google.com Orlando Andico +63.2.976.8659 | +63.920.903.0335 _________________________________________________ Philippine Linux Users' Group (PLUG) Mailing List http://lists.linux.org.ph/mailman/listinfo/plug Searchable Archives: http://archives.free.net.ph
