Presentation 2005-01-21
An Extended Range Perturbed Correlation Method and Experimental Evaluation : An Improved Correlation Method for Character Recognition
Koji KITAMURA, Marie NIKAIDO, Yumi NAKASHIMA, Michio YASUDA,
PDF Download Page PDF download Page Link
Abstract(in Japanese) (See Japanese page)
Abstract(in English) When calculate the similarity between standard character and unknown character, it is well known that maximize the parameter of two dimension affine transformation is effective in the character recognition technique for using the correlation method. Before, the amount of the purturbation done only ±1 unit about each parameter for an affine transformation, we expand it up to ±3 unit. Actually, the simulation experiment marks almost 99 % correct reading rate whithout feature extraction that used 10 kind of the figure 13,830 samples of Electrotechnical Laboratory handwritten numeral character database et16.
Keyword(in Japanese) (See Japanese page)
Keyword(in English) Correlation / Perturbation / Pattern matching / Handwritten Numeral Character
Paper # PRMU2004-163
Date of Issue

Conference Information
Committee PRMU
Conference Date 2005/1/14(1days)
Place (in Japanese) (See Japanese page)
Place (in English)
Topics (in Japanese) (See Japanese page)
Topics (in English)
Chair
Vice Chair
Secretary
Assistant

Paper Information
Registration To Pattern Recognition and Media Understanding (PRMU)
Language JPN
Title (in Japanese) (See Japanese page)
Sub Title (in Japanese) (See Japanese page)
Title (in English) An Extended Range Perturbed Correlation Method and Experimental Evaluation : An Improved Correlation Method for Character Recognition
Sub Title (in English)
Keyword(1) Correlation
Keyword(2) Perturbation
Keyword(3) Pattern matching
Keyword(4) Handwritten Numeral Character
1st Author's Name Koji KITAMURA
1st Author's Affiliation Faculty of Infomatics, Meisei University()
2nd Author's Name Marie NIKAIDO
2nd Author's Affiliation Faculty of Infomatics, Meisei University
3rd Author's Name Yumi NAKASHIMA
3rd Author's Affiliation Faculty of Infomatics, Meisei University
4th Author's Name Michio YASUDA
4th Author's Affiliation Faculty of Infomatics, Meisei University
Date 2005-01-21
Paper # PRMU2004-163
Volume (vol) vol.104
Number (no) 573
Page pp.pp.-
#Pages 6
Date of Issue