| 摘要 |
When an electronic document is to be generated from an image of a document, reproduction of both an appearance of the document image and a logical structure is difficult because of format limitations imposed on the electronic document. According to the present invention, from among column arrangements allowed by the specifications of the electronic document, a column arrangement is selected by using which the positions of character areas in an input image and the logical structure analyzed from the input image can be most appropriately reproduced, and is used to generate an electronic document. |
