| 摘要 |
<p>A character-detecting section detects characters from an image that has been read from a document. In order to allow a character-recognizing section to make a judgement as to whether or not the character is identical to any character of specific documents, such as paper money and securities, stored in a RAM, a discrimination-level setting section generates discrimination patterns corresponding to the discrimination level, and supplies them to the character-recognizing section. As a result of the judgement based on the discrimination at a low level, if the document is judged as one of the specific documents, a specific-document judging section controls the discrimination-level setting section in order to allow the character-recognizing section to make a re-judgement for identification. Thus, the discrimination-level setting section supplies discrimination patterns at a higher level to the character-recognizing section. This arrangement makes it possible to shorten the processing time that is required for discrimination as to whether or not the document in question is any of the specific documents, as well as making it possible to ensure the reliability of discrimination. <IMAGE></p> |
