Evaluation and Synthesis of Feature Vectors for Handwritten Numeral Recognition

Fumitaka KIMURA; Shuji NISHIKAWA; Tetsushi WAKABAYASHI; Yasuji MIYAKE; Toshio TSUTSUMIDA

Evaluation and Synthesis of Feature Vectors for Handwritten Numeral Recognition

Fumitaka KIMURA, Shuji NISHIKAWA, Tetsushi WAKABAYASHI, Yasuji MIYAKE, Toshio TSUTSUMIDA

Full Text Views

0

Cite this

Summary :

This paper consists of two parts. The first part is devoted to comparative study on handwritten ZIP code numeral recognition using seventeen typical feature vectors and seven statistical classifiers. This part is the counterpart of the sister paper Handwritten Postal Code Recognition by Neural Network - A Comparative Study" in this special issue. In the second part, a procedure for feature synthesis from the original feature vectors is studied. In order to reduce the dimensionality of the synthesized feature vector, the effect of the dimension reduction on classification accuracy is examined. The best synthesized feature vector of size 400 achieves remarkably higher recognition accuracy than any of the original feature vectors in recognition experiment using a large number of numeral samples collected from real postal ZIP codes.

Publication: IEICE TRANSACTIONS on Information Vol.E79-D No.5 pp.436-442

Publication Date: 1996/05/25

Publicized

Online ISSN

DOI

Type of Manuscript: Special Section PAPER (Special Issue on Character Recognition and Document Understanding)

Category: Comparative Study

Cite this

Copy

Fumitaka KIMURA, Shuji NISHIKAWA, Tetsushi WAKABAYASHI, Yasuji MIYAKE, Toshio TSUTSUMIDA, "Evaluation and Synthesis of Feature Vectors for Handwritten Numeral Recognition" in IEICE TRANSACTIONS on Information, vol. E79-D, no. 5, pp. 436-442, May 1996, doi: .
Abstract: This paper consists of two parts. The first part is devoted to comparative study on handwritten ZIP code numeral recognition using seventeen typical feature vectors and seven statistical classifiers. This part is the counterpart of the sister paper Handwritten Postal Code Recognition by Neural Network - A Comparative Study" in this special issue. In the second part, a procedure for feature synthesis from the original feature vectors is studied. In order to reduce the dimensionality of the synthesized feature vector, the effect of the dimension reduction on classification accuracy is examined. The best synthesized feature vector of size 400 achieves remarkably higher recognition accuracy than any of the original feature vectors in recognition experiment using a large number of numeral samples collected from real postal ZIP codes.
URL: https://global.ieice.org/en_transactions/information/10.1587/e79-d_5_436/_p

Copy

@ARTICLE{e79-d_5_436,
author={Fumitaka KIMURA, Shuji NISHIKAWA, Tetsushi WAKABAYASHI, Yasuji MIYAKE, Toshio TSUTSUMIDA, },
journal={IEICE TRANSACTIONS on Information},
title={Evaluation and Synthesis of Feature Vectors for Handwritten Numeral Recognition},
year={1996},
volume={E79-D},
number={5},
pages={436-442},
abstract={This paper consists of two parts. The first part is devoted to comparative study on handwritten ZIP code numeral recognition using seventeen typical feature vectors and seven statistical classifiers. This part is the counterpart of the sister paper Handwritten Postal Code Recognition by Neural Network - A Comparative Study" in this special issue. In the second part, a procedure for feature synthesis from the original feature vectors is studied. In order to reduce the dimensionality of the synthesized feature vector, the effect of the dimension reduction on classification accuracy is examined. The best synthesized feature vector of size 400 achieves remarkably higher recognition accuracy than any of the original feature vectors in recognition experiment using a large number of numeral samples collected from real postal ZIP codes.},
keywords={},
doi={},
ISSN={},
month={May},}

Copy

TY - JOUR
TI - Evaluation and Synthesis of Feature Vectors for Handwritten Numeral Recognition
T2 - IEICE TRANSACTIONS on Information
SP - 436
EP - 442
AU - Fumitaka KIMURA
AU - Shuji NISHIKAWA
AU - Tetsushi WAKABAYASHI
AU - Yasuji MIYAKE
AU - Toshio TSUTSUMIDA
PY - 1996
DO -
JO - IEICE TRANSACTIONS on Information
SN -
VL - E79-D
IS - 5
JA - IEICE TRANSACTIONS on Information
Y1 - May 1996
AB - This paper consists of two parts. The first part is devoted to comparative study on handwritten ZIP code numeral recognition using seventeen typical feature vectors and seven statistical classifiers. This part is the counterpart of the sister paper Handwritten Postal Code Recognition by Neural Network - A Comparative Study" in this special issue. In the second part, a procedure for feature synthesis from the original feature vectors is studied. In order to reduce the dimensionality of the synthesized feature vector, the effect of the dimension reduction on classification accuracy is examined. The best synthesized feature vector of size 400 achieves remarkably higher recognition accuracy than any of the original feature vectors in recognition experiment using a large number of numeral samples collected from real postal ZIP codes.
ER -