发明名称 DOCUMENT READING DEVICE
摘要 PURPOSE:To enable high-speed extraction of character string by dividing picture elements of reading area of a document into specified number of small areas and providing a picture compressing device, a characteristic extracting device, an area detecting device and a character string detecting device. CONSTITUTION:A read picture of a document stored in an image information storing section 12 is divided into small areas of specified number of units. In the small areas, pictures continuing at specified intervals is extracted by a compressed picture processing section 15. Areas in which a picture is connected at a specified distance or more is detected by an area detecting section 16. The detection and extraction of character string pictures are made in a character string detecting section 17 from the shape of the area detected by the detecting section 16 and information of picture element characteristic picture that indicates the presence of a lateral straight line, and supplied to a character recognizing section 19. Thus, each character in the character string picture is cut out and recognized in the recognizing section 19. By compressing the read picture first and detecting a character area by using the compressed picture, the character string contained in a document can be extracted at high speed.
申请公布号 JPS61175875(A) 申请公布日期 1986.08.07
申请号 JP19850017266 申请日期 1985.01.31
申请人 MITSUBISHI ELECTRIC CORP 发明人 MAEDA YOJI;NANBU HAJIME
分类号 G06K9/20;G06K9/36 主分类号 G06K9/20
代理机构 代理人
主权项
地址