ICR, MICR and OMR
While the accurate recognition of Latin-script typewritten text in large variety of fonts is
now considered largely a solved problem, recognition of hand printing and handwriting in
general are still the subject of active research. Recognition of
scanned handwritten text is
known as ICR which stands for Intelligent Character Recognition. ICR is an extension of
OCR which explicitly includes handwritten characters. There are a large number of
research efforts in this field and does not seem to be ending anytime soon. Most working
testbeds work on limited vocabulary and clean handwriting. A robust mechanism does not
exist to perform this task effectively.
Errors often consist of misreading one character for
another. For example:
1. 0 (zero) for O (oh)
2. k for lc
3. m for nn.
Higher rates of recognition is made possible through the use of contextual or grammatical
information. For example, recognizing entire words from a dictionary is easier than trying
to parse individual characters from script. Reading the Amount line of a check (where
you know the information should be a written out number) is an example of a smaller
dictionary where accuracy rates can be increased greatly. Knowledge of the grammar of
the language being scanned can also help determine if a word is likely to be a verb or a
noun, for example, allowing greater accuracy. For example, using a lexicon, an ICR
system will deduce that the word "tke" at the beginning of a sentence is a mistake and
can potentially be replaced with the word "the".
Another offshoot of OCR is OMR. OMR stands for Optical Mark Recognition. OMR is used
to read forms which have variable size field entries and the entry can be in form of a
check mark, a cross, or some scribble. These forms include but are not
automated entry sheets where the user needs to fill in an oval shape with a number 2
pencil. OMR usually needs a blank form and a definition of entry zones, then the OMR
detects marks in the scaned image and classifies them accordingly.
An example application of OMR technology is in the use
of OMR to automatically check
the examination answer sheets of objective-type
questions in University entrance
Magnetic Ink Character Recognition, or MICR, is a special kind of optical character
recognition technology that was adopted mainly by the financial or banking industry to
facilitate the processing of checks. Almost all checks include MICR characters
bottom of the cheque.
The general problem of character recognition is limited in its usefulness to the degree of
the accuracy of the recognition. Through the use of specially designed characters printed
with ink with a magnetic signature, the error rate for scanning the
numbers at the bottom
of a typical check is smaller than with usual OCR systems. Furthermore, the letters can
be recognized with a device similar in nature to the head of an audio tape recorder. The
reliability of MICR recognition is significantly higher than
when processed by hand, and is
therefore highly useful. Error rates typically range around one read error for every 20,000 to 30,000 checks.
MICR can be added to checks with existing print technology using the iron-oxide inks for
almost no additional cost. Today laser printer toner cartridges are available with MICR ink for printing your own.