Method and means for improving optical character recognition (OCR) of printed documents

摘要

The document markers containing the first values, which are dependent on the layout and content of the document and are assigned by the creation or processing software, are provided as machine readable symbolic representations on the document surface in the printed form. Markers contain coded document placement information and values assigned on a sequence of original text, which values include decimation sequences, error correction codes, or checksums depending on the text. When scanning with optical character recognition or when performing other digitized reproduction, the markers are also scanned. The scanning computer has corresponding software and allocates second values depending on the arrangement and content of the reproduced document. When comparing the first and second decimation sequences, line and character errors are detected and some errors are corrected to produce rearranged sequences. An optional correction code may provide a better correction function when applied to the rearranged reproduced document sequences and an optional check-sum comparison may be used to verify that the accuracy of the reproduced sequences is correct.