| abstract |
A character string extraction device according to the present invention includes a replacement information registering unit in which replacement information to replace character information expected to be erroneously recognized is registered, a candidate character data registering unit in which supposed candidate character data is registered, an image information converting unit for converting the read image information into character information, a character information replacing unit for replacing a specific character with a designated character when the character information includes the specific character and generating read character data by using the converted character information when the character information does not include the specific character, a search character generating unit for replacing a predetermined character of the read character data with a special character and, generating search character data from the read character data, and a first comparing unit for comparing the search character data with the candidate character data. |