Presentation 1998/12/17
Handwritten Similar Characters Recognition by the Unified Neural Network Attention to Peripheral Depth
Keitaro Hori, Hiroki Sugawara, Akiyoshi Itoh,
PDF Download Page PDF download Page Link
Abstract(in Japanese) (See Japanese page)
Abstract(in English) We have propsed the feature extraction method : peripheral local outline vector(P-LOVE)already. P-LOVE extracts a feature at a peripheral point on a character which is scanning from 8 directions around the character. In this paper, we propose the unified neural network(NN)system in which each NN clssifies patterns depending on the P-LOVE featrues for the different peripheral depth. Experimental results show taht the system achieves higher recognition rates than a single NN for handwritten similar characters of ETL9B. In addition, we consider a rejecting method by using a majority decision of NNs in the 1st stage.
Keyword(in Japanese) (See Japanese page)
Keyword(in English) character recognition / similar characters / neural network / peripheral depth / peripheral local outline vector / ETL9B
Paper # PRMU98-145
Date of Issue

Conference Information
Committee PRMU
Conference Date 1998/12/17(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) Handwritten Similar Characters Recognition by the Unified Neural Network Attention to Peripheral Depth
Sub Title (in English)
Keyword(1) character recognition
Keyword(2) similar characters
Keyword(3) neural network
Keyword(4) peripheral depth
Keyword(5) peripheral local outline vector
Keyword(6) ETL9B
1st Author's Name Keitaro Hori
1st Author's Affiliation NIHON Univ.Graduate School()
2nd Author's Name Hiroki Sugawara
2nd Author's Affiliation NIHON Univ.Graduate School
3rd Author's Name Akiyoshi Itoh
3rd Author's Affiliation NIHON Univ.College of Sci.&Tech., Dep.of Electronic Engineering
Date 1998/12/17
Paper # PRMU98-145
Volume (vol) vol.98
Number (no) 489
Page pp.pp.-
#Pages 8
Date of Issue