摘要 |
PURPOSE:To read out characters in a document without requiring a sufficient clear area by comparing an unentered document image with an entered document image and extracting a character image entered in a character entering frame. CONSTITUTION:An unentered document out of documents whose character entering frames are printed out with readable color is read out by a scanning part 11 and the image of the unentered document is stored in a magnetic disk device 12. The unentered document image corresponding to an entered document image obtained by reading out a character-entered document is extracted from a magnetic disk device 12, both the images are compared with each other and the character image entered in the character frame is extracted. The extracted character image is recognized by a recognizing part 14.
|