OCR-on-a-Chip is the most
advanced breakthrough in OCR technology since Ligature's introduction
of neural networks to OCR in 1991.
OCR-on-a-Chip is a powerful OCR tool that provides reading capabilities
to any machine or integrated system without the need for a PC or
extensive memory. OCR, previously a specialized document imaging
tool, will now be available for use in industries such as robotics,
digital cameras, heavy industry, manufacturing and consumer product
development. The market potential for this development is vast and,
as such, eludes limiting definitions. Ligature is currently discussing
development projects with several companies in a variety of industries.
The Development of OCR-on-a-Chip
The objective of OCR-on-a-Chip is to enable integrated systems or
digital cameras that are faced with cost, space and power limitations,
to achieve advanced text reading ability. Developing this solution
essentially involved downscaling the memory requirements and hardware
price needed, in order to achieve high power text recognition without
compromising speed or accuracy.
The OCR-on-a-Chip solution allows Ligature partners to take OCR
technology one step further. With OCR-on-a-Chip you can not only
process full pages of text on a desktop machine, but can also "read"
short captions or clips of text using a hand-held gadget or digital
camera. The application opportunities for using OCR-on-a-Chip are
numerous and vary from scanning a word from a document using a pen
scanner, to clipping a license plate from a car photo taken with
a digital camera. The uniqueness of OCR-on-a-Chip is that it runs
workstation-type intelligence on a low-powered, inexpensive, easily
glued on chip.
The requirements for OCR-on-a-Chip seem incredible when compared
to those for traditional OCR packages. Ligature's stand-alone OCR
application, CharacterEyes, known for its small memory footprint,
requires less memory than any of its competitors with a required
8 MB RAM and 4 MB hard disk space on a PC. In contrast, OCR-on-a-Chip
only requires a CPU,
500K - 1100K ROM, and a mere 64K of RAM.
Opportunities Abound
Leaders from a wide spectrum of industries have voiced tremendous
enthusiasm for the new development, and applications are expected
to be developed on a joint basis. The first product on the
market featuring Ligature's OCR-on-a-Chip technology is Wizcom's
Quicktionary, winner of the Best of Byte award at CeBIT.
Quicktionary is a hand-held pen-scanner that uses 64K RAM
to scan and translate text clips. Ligature is now actively
seeking additional partners to develop applications based
on OCR-on-a-Chip. Dynamic companies interested in joint development
projects should write to
isales@ligatureltd.com
