摘要 |
PROBLEM TO BE SOLVED: To save a time and labor for generating format information for each document in a method for recognizing, by using an optical character reading device, a mark for specifying a character to which a mark is added. SOLUTION: The picture of a document 1 inputted through an optical sensor 21 is stored in an image memory 23. A mark field segmenting part 25 segments the field of a character unit. According to the size of the character of the field, a mark recognizing part 26 detects, and a character recognizing part 27 operates the character recognition of a character to which a mark is not added. Then, the character to which the mark 'woman' is added is specified, regardless of the order, based on a mark content 28b. |