Presentation 1998/2/19
Fast, Precise Pre-classification Method Using Projections of Feature Regions
Katsuhito Fujimoto, Hiroshi Kamada, Koji Kurokawa,
PDF Download Page PDF download Page Link
Abstract(in Japanese) (See Japanese page)
Abstract(in English) We propose a fast, precise pre-classification method for a lot of categories. Conventional pre-classification uses simpler discrimination functions than those used by detail-classification. But sufficient simplification for matching acceleration sacrifices some degree of matching accuracy. On the other hand, the proposed method can preserve the accuracy of the detail-classification and achieve drastic acceleration of matching process by holding a mapping from low-dimensional reference features to candidate category sets by estimating projections of each category's feature regions used in the detail-classification. The proposed method applied to Japanese character recognition can accelerate matching process by 3 times over the conventional method with no decrease of matching accuracy, which confirms the effectiveness of the proposed method.
Keyword(in Japanese) (See Japanese page)
Keyword(in English) pattern recognition / matching / pre-classification / character recognition / feature region / projection
Paper #
Date of Issue

Conference Information
Committee PRMU
Conference Date 1998/2/19(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) Fast, Precise Pre-classification Method Using Projections of Feature Regions
Sub Title (in English)
Keyword(1) pattern recognition
Keyword(2) matching
Keyword(3) pre-classification
Keyword(4) character recognition
Keyword(5) feature region
Keyword(6) projection
1st Author's Name Katsuhito Fujimoto
1st Author's Affiliation FUJITSU LABORATORIES LTD. Multimedia Systems Laboratories()
2nd Author's Name Hiroshi Kamada
2nd Author's Affiliation FUJITSU LABORATORIES LTD. Multimedia Systems Laboratories
3rd Author's Name Koji Kurokawa
3rd Author's Affiliation FUJITSU LABORATORIES LTD. Multimedia Systems Laboratories
Date 1998/2/19
Paper #
Volume (vol) vol.97
Number (no) 558
Page pp.pp.-
#Pages 8
Date of Issue