| 摘要 |
PROBLEM TO BE SOLVED: To provide an image processing method which can easily retrieve a file. SOLUTION: When an input image is filed into a PDF file, character areas and photographic areas of the input image are discriminated, and image feature quantities such as the quantity of ruled lines, a character size, the number of character colors, and the quantity of dots are extracted per discriminated area, and attribute information of the input image is generated on the basis of extracted image feature quantities, in concrete terms, a goodness-of-fit (%) indicating which kind of images of a table, a document, a presentation, a magazine, etc. the input image belongs to with high probabilities is calculated, and this attribute information is added to the PDF file. COPYRIGHT: (C)2006,JPO&NCIPI
|
