摘要 |
PROBLEM TO BE SOLVED: To eliminate the erroneous extraction or no extraction of character lines in the processing to cope with a sub-formated document or a thinning document to be assumed by analyzing data, which are considered as character components, by a format extracting device, several times while changing discrimination conditions. SOLUTION: A reading system performs format extraction 12 and character recognition 13 from the binarized image of a document inputted by an image scanner 11. In the format extraction 12, the detailed processing of character line extraction corresponding to the sub-formated slip or the thinning document is performed at several stages while changing the discrimination conditions based on a reading area on the slip and description rules. The information of frames/character lines extracted by the format extraction 12 and the character information-read result of character codes extracted by the character recognition 13 are stored as output data 14 by a user system. Based on the output data 14, the user system performs user confirmation/correction 15. Thus, the erroneous extraction or no extraction of the sub-formated slip or the thinning document is eliminated.
|