Presentation 2004/11/12
Discrimination of signboard pictures using a mobile camera
Takayasu YAMAGUCHI, Hiroshi AONO, Sadayuki HONGO,
PDF Download Page PDF download Page Link
Abstract(in Japanese) (See Japanese page)
Abstract(in English) In this paper we evaluate some learning and discrimination methods for discrimination of signboard pictures from viewpoint of use in mobile environment. Fisher's Linear Discriminant (FLD), the subspace method, the projection distance method, the linear vector quantization, the support vector machine (SVM), the naive Bayes (NB) and the k-nearest neighbor method (k-NN), which are major traditional methods, are evaluated from the time for learning and discrimination and the score which corresponds the correct discrimination rate. Our experiments show that scores of the SVM, the projection distance method, the subspace method, the k-NN, the NB and the FLD with the dimension compression in the combination of every two classes are high, and also show that the time for learning and discrimination in k-NN and NB is short. Finally, we conclude that the NB has the advantage of small increase in the discrimination time over increase in the number of registered signboard against the k-NN.
Keyword(in Japanese) (See Japanese page)
Keyword(in English) Mobile camera / Signboard / Naive Bayes / k-Nearest Neighbor
Paper # PRMU2004-106,HIP2004-46
Date of Issue

Conference Information
Committee HIP
Conference Date 2004/11/12(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 Human Information Processing (HIP)
Language JPN
Title (in Japanese) (See Japanese page)
Sub Title (in Japanese) (See Japanese page)
Title (in English) Discrimination of signboard pictures using a mobile camera
Sub Title (in English)
Keyword(1) Mobile camera
Keyword(2) Signboard
Keyword(3) Naive Bayes
Keyword(4) k-Nearest Neighbor
1st Author's Name Takayasu YAMAGUCHI
1st Author's Affiliation Network Management Development Department. NTT DoCoMo, Inc.()
2nd Author's Name Hiroshi AONO
2nd Author's Affiliation Network Management Development Department. NTT DoCoMo, Inc.
3rd Author's Name Sadayuki HONGO
3rd Author's Affiliation Network Management Development Department. NTT DoCoMo, Inc.
Date 2004/11/12
Paper # PRMU2004-106,HIP2004-46
Volume (vol) vol.104
Number (no) 450
Page pp.pp.-
#Pages 6
Date of Issue