caa a22 4500 467884161 CHVBK 20180406152729.0 cr unu---uuuuu 170328e20060601xx s 000 0 eng 10.1007/s10032-005-0003-8 doi (NATIONALLICENCE)springer-10.1007/s10032-005-0003-8 Fourier-Mellin registration of line-delineated tabular document images [Elektronische Daten] [Luke Hutchison, William Barrett] Image registration (or alignment) is a useful preprocessing tool for assisting in manual data extraction from handwritten forms, as well as for preparing documents for batch OCR of specific page regions. A new technique is presented for fast registration of lined tabular document images in the presence of a global affine transformation, using the Discrete Fourier--Mellin Transform (DFMT). Each component of the affine transform is handled separately, which dramatically reduces the total parameter space of the problem. This method is robust and deals with all components of the affine transform in a uniform way by working in the frequency domain. The DFMT is extended to handle shear, which can approximate a small amount of perspective distortion. In order to limit registration to foreground pixels only, and to eliminate Fourier edge effects, a novel, locally adaptive foreground-background segmentation algorithm is introduced, based on the median filter, which eliminates the need for Blackman windowing as usually required by DFMT image registration. A novel information-theoretic optimization of the median filter is presented. An original method is demonstrated for automatically obtaining blank document templates from a set of registered document images. Springer-Verlag, 2005 Document image registration nationallicence Deskewing nationallicence Background removal nationallicence Fourier-Mellin transform nationallicence Transformation reversal nationallicence Locally-adaptive thresholding nationallicence Microfilm processing nationallicence Tabular document images nationallicence Family history technology nationallicence Hutchison Luke Department of Computer Science, Brigham Young University, Provo, UT, USA aut Barrett William Department of Computer Science, Brigham Young University, Provo, UT, USA aut International Journal of Document Analysis and Recognition (IJDAR) Springer-Verlag 8/2-3(2006-06-01), 87-110 1433-2833 8:2-3<87 2006 8 10032 https://doi.org/10.1007/s10032-005-0003-8 text/html Onlinezugriff via DOI 1 research-article jats NATIONALLICENCE 856 40 https://doi.org/10.1007/s10032-005-0003-8 text/html Onlinezugriff via DOI NATIONALLICENCE 700 1- Hutchison Luke Department of Computer Science, Brigham Young University, Provo, UT, USA aut NATIONALLICENCE 700 1- Barrett William Department of Computer Science, Brigham Young University, Provo, UT, USA aut NATIONALLICENCE 773 0- International Journal of Document Analysis and Recognition (IJDAR) Springer-Verlag 8/2-3(2006-06-01), 87-110 1433-2833 8:2-3<87 2006 8 10032 Metadata rights reserved Springer special CC-BY-NC licence nationallicence BK010053 XK010053 XK010000 NATIONALLICENCE NATIONALLICENCE NL-springer