Title :
A form dropout method based on line-elimination and image-subtraction
Author :
Shima, Yoshihiro ; Ohya, Hiroshi ; Yasuda, Michio
Author_Institution :
Fac. of Phys. Sci. & Eng., Meisei Univ., Tokyo, Japan
fDate :
29 Aug.-1 Sept. 2005
Abstract :
A method of preprint elimination for form images is proposed. Form images have fixed parts, such as character strings and lines. These fixed parts are known as preprint, which are not used for data entry. This dropout method is based on image subtraction and line elimination for distorted images. The location, rotation and magnification are modified for distorted form images. Character patterns and short ruled lines are eliminated by subtraction of bitmap template images. Long ruled lines are extracted and directly eliminated by using run data.
Keywords :
business forms; data compression; document image processing; bitmap template image; character line elimination; character pattern; character string; form dropout method; form image preprint elimination; image distortion; image-subtraction; Banking; Data compression; Data mining; Government; Image coding; Image processing; Informatics; Insurance; Optical character recognition software; Pixel;
Conference_Titel :
Document Analysis and Recognition, 2005. Proceedings. Eighth International Conference on
Print_ISBN :
0-7695-2420-6
DOI :
10.1109/ICDAR.2005.9