Optical character recognition (OCR) systems convert digital images of text into symbolic strings (for example, ASCII). Digital images of text are typically created by scanning paper documents containing printed or handwritten text, from video images of scenes with text, or by digital pens. The OCR-generated symbolic form of the paper documents allows multiple users to simultaneously access, search, edit, and transmit the contents of these pages. In fact, with the help of machine translation and speech synthesis systems, the information in paper documents can now be accessed by people across language boundaries and by those with visual impairment.