| 摘要 |
PROBLEM TO BE SOLVED: To easily discriminate the content of a retrieved document file. SOLUTION: A document delivery server 11 acquires a document file 17 corresponding to a retrieval request for a document file 17 received from a cellular phone 12 from a file server 13. Respective document files 17 are imaged to generate a document image 18. A feature value is extracted from each page image 18a of the document image 18, and based on the feature value extraction result, it is determined whether the document image 18 is a text document image. If the document image 18 is not a text document image, a thumbnail image 42 is generated based on a page image 18a which is most characteristic among the page images 18a. If the document image 18 is a text document image, an extraction text 43 is extracted from a top page of an original document file 17. The thumbnail image 42 and the extraction text 43 are sent to the cellular phone 12 as retrieval results. COPYRIGHT: (C)2011,JPO&INPIT |
