ICR, or Intelligent Character Recognition, is an advanced form of OCR that is able to recognize handwritten text as well as unusual fonts, in addition to standard typewritten text. The development of ICR has added flexibility to the realm of OCR, because software no longer depends on the conformity of text to strict standards for the shape of characters. As ICR technology develops, more and more processes can become automated, since human involvement is no longer necessary to interpret handwritten characters.
ICR is especially useful for automated forms processing, which allows standardized forms to be handled by machine. Forms processing software is able to extract text from specified fields on a form and enter in into a computer database. Before ICR was developed, this process could only be automated for printed forms, while handwritten forms were processed manually. With the development and improvement of ICR, this tedious, repetitive process no longer requires continuous attention.
ICR is used in conjunction with OCR across a wide span of industries, because many companies and organizations require people to fill out forms by hand. Because many large offices process enormous volumes of paper forms daily, it is impractical to have employees retype the responses in each form. Instead, the papers can simply be scanned, and processed using ICR, just as other documents are processed with OCR tools.
For now, ICR does not have the same accuracy levels that OCR has attained. While modern OCR technology is near-perfect, ICR often runs into trouble when it encounters difficult-to-read characters. Even so, it is possible to achieve accuracy rates of 97% or higher, especially if several different ICR engines are used. Another sub-field of OCR, called IWR, or Intelligent Word Recognition, attempts to take ICR to an even higher level. IWR recognizes cursive text in addition to printed handwriting, to provide maximum flexibility in document automation. However, this area still requires a great deal of research, due to the difficulty of interpreting handwriting styles that vary greatly from person to person. If IWR is developed to become more accurate, it can help tremendously with handling those documents that currently can only be done by hand.
ICR is one of the aspects of OCR with the most room for development. At present, many OCR solutions incorporate ICR capabilities to aid the document automation process. One such solution is SoftWorks AI’s Trapeze, which takes care of all types of forms processing. Trapeze contains ICR in order to provide users with the most comprehensive services available.