| 摘要 |
<P>PROBLEM TO BE SOLVED: To reduce an image information amount required for character recognition to shorten processing time for the character recognition by making it unnecessary to use a medium of fixed form format when character information described on the medium is fetched. <P>SOLUTION: A CPU 1 specifies an area with thickness different from that of an original on the whole surface of the original based on the thickness of the original detected by an original thickness detection sensor 25 before reading an image of the original by a scanner 3, and before the original to be conveyed by an ADF 2 reaches a reading position of the scanner 3, and makes the scanner 3 read an image only in the specified area on the original. Image recognition is performed to image data of the read image by a character recognition processing part 6 to obtain the recognized character information. In addition, when a plurality of areas are specified, the processing content such as stop of reading, selection of an area with the highest similarity to a default area, and selection of an area by a user is selected and executed. <P>COPYRIGHT: (C)2011,JPO&INPIT |
