| 摘要 |
A plurality of background texture patterns each of which includes a region not having a dot and having a feature different from those of the other background texture patterns, are combined, and information is embedded in a background region of a document image. Then, a mathematical-morphology-based conversion is performed on the document image, and thereby a feature of a region not having a dot in each of the background texture patterns is extracted in order to identify the embedded information.
|
