| 摘要 |
PROBLEM TO BE SOLVED: To perform an optimum recognition processing by incorporating a plurality of OCR engines and combining/selecting an optimum OCR engine according to the image state of a target to be recognized. SOLUTION: A recognition image read out by a image reader which is a reading means, a target recognition image in a document file such as an e-mail, a file attached to an e-mail and a facsimile, or the like is acquired by an acquisition means (S21). In accordance with processing to be applied to the recognition image and the state of the recognition image acquired by the acquisition means, whether an OCR engine selected from a plurality of OCR engines is set or not is confirmed (S22), and when the OCR engine is set (Yes in S22), recognition processing is performed by using the specified OCR engine (S23). When the OCR engine is not set in the processing S22 (No in S22), recognition processing using an OCR engine previously set in the apparatus is performed (S24). COPYRIGHT: (C)2007,JPO&INPIT
|
