| 摘要 |
The method involves acquiring a digital image (19) of a document (2), and processing the acquired digital image. The digital image is processed by detection of boxes (8, 14) on the acquired image by template matching, where each box is detected individually by matching with a box template, and each box template represents a fraction of the box. The detection of the boxes is carried out by reproducing physical deformations of the document during acquisition, without digital geometrical correction of the physical deformations of the document. An independent claim is also included for an optical reading device for reading a document. |
