| 摘要 |
Scanning and processing a multi-page document is described. Scan images are generated by optically scanning a sequence of pages of a multi-page document such as a book. The scan images are processed to generate page images corresponding to original pages of the multi-page document, e.g. by deskewing and removing black border areas. After generating a proposed page image it is checked if there is a chance that that image contains errors, by detecting if its image parameters, such as the text area coordinates (X1,X2,X3,X4,Y1,Y2) are according to target criteria which are based on a statistical analysis of the document. If the page image satisfies the target criteria, the page image is automatically accepted. However, if it does not, the page image (65,66) is displayed for an operator to accept or to adjust the page image.
|
