JPH05347698

PURPOSE: To effectively detect a highlight region by scanning a document including the highlight region using a gray scale scanner, separating it into a highlight part and a non-highlight part and extracting a text from the highlight region using an optical character recognition technique. CONSTITUTION: A system includes the scanner 5 for digitizing the document in pixel unit and supplying the resultant data structure. This scanner preferably supplies a gray scale picture. This data structure is separated into the highlight part and the non-highlight part by the gray scale scanner, and one or plural positions of mark are identified using the gray scale picture. This mark is enclosed with a region having a non-white gray scale different from a gray scale. Then, this non-white gray scale region is included with an almost solid region. Then, the data is transmitted to an application particular hardware 13 by the optical character recognition technique.