Optical Character Recognition is the science of capturing characters from documents and translating them into data that can be used within computer systems.
OCR Systems require an optical reading device such as a Scanner to capture characters. Depending on the type of document and the scanning equipment used, data can be captured and processed immediately as the paper passes through the scanning equipment or passed to sophisticated software that can carry out the recognition from the image as a secondary process.
In reality, most OCR systems utilise both hardware and software recognition to provide the most satisfactory results on machine printed text, barcode, E13B or handwritten characters.
To reduce the ambiguities between the shapes of similar characters (such as B and 8, Z and 2) a special OCR-B font was developed and is now widely used within the Business community for printing invoice details on payment slips and remittances.
An increasing number of Organisations are now aware of the extensive possibilities for OCR systems as almost any document can be scanned and relevant information extracted, recognised and stored.
Numerous Companies have automated their payment processing departments by using documents printed in the OCR-B font (which has become the industry standard) and bringing in OCR Scanning systems to improve efficiency.
Whilst excellent results can be achieved with OCR using matrix matching, feature analysis and neural network techniques, a complete document processing system will require additional facilities to correct unrecognised characters and carry out validation, reporting and output conversion routines.