|
|
|
Image Recognition Technology
|
|
|
|
|
|
Navigate the Document Imaging Wiki through the navigation at the top of the page,
or simply type in the term you are looking for in the search pane.
OCR/ICR — (Optical Character Recognition/Intelligent Character Recognition). The use of technology to identify characters on a piece of paper. Once identified, these characters can be translated into a computer language, such as ASCII or XML. OCR is often referred to as applying to machine-printed text only, and ICR to hand-printed text, although technically, these differentiations on not correct.
IDR — (Intelligent Document Recognition). This is the next-generation of OCR/ICR technology. Instead of just recognizing characters, like OCR/ICR, IDR adds some “intelligence” to the recognition process, which enables it to identify words and their context. Popular IDR applications include invoice capture and the digital mailroom.
OMR — (Optical Mark Recognition). Using technology to identify shaded areas on a page. Analog OMR, which is most widely used for test scoring, utilizes infra-red light source top dropout all colors on a page, except for those created generated with a carbon-based pencil or carbon-colored ink. Recently, image-based OMR techniques have been developed, including MDO (Multi-Color Dropout) technology by Kodak.
NHR — Natural handwriting recognition; technology for enabling automated recognition of cursive and hand-printed characters. To be most effective, NHR technology is initially applied on a character-level, with the results then being taken in context with the characters they are connected to (in the same word, for example) and compared to some form of dictionary of terms. This could be a dictionary of addresses, for example, in a postal sorting application.
Major Image Recognition vendors include: A2ia, ABBYY, I.R.I.S., Mitek, Nuance, Oce-ODT, and Parascript.
Favorites
My Settings
Help
Sign Out
Sign in