Optical character recognition, usually abbreviated to OCR, is the translation of:
All the open source software are based on the tesseract OCR engine.
tesseract imagename outputbase [-l lang] [--oem ocrenginemode] [--psm pagesegmode] [configfiles...]
VietOCR is the only one that I found which:
For windows, you have also FreeOCR 2.6. It work well but you can't process more than one page at a time. May be it's the good way because you always need to clean the result.