Off-line, handwritten numeral recognition by perturbation method

被引：115

作者：

Ha, TM

Bunke, H

机构：

[1] University of Berne, Institute of Computer Science and Applied Mathematics, Neubrueckstr W

来源：

IEEE TRANSACTIONS ON PATTERN ANALYSIS AND MACHINE INTELLIGENCE | 1997年 / 19卷 / 05期

关键词：

writing habits and styles; writing instruments; reversing process; perturbation method; decision combination; k-nearest neighbor rule; neural networks;

D O I：

10.1109/34.589216

中图分类号：

TP18 [人工智能理论];

学科分类号：

081104 ; 0812 ; 0835 ; 1405 ;

摘要：

This paper presents a new approach to off-line, handwritten numeral recognition. From the concept of perturbation due to writing habits and instruments, we propose a recognition method which is able to account for a variety of distortions due to eccentric handwriting. We tested our method on two worldwide standard databases of isolated numerals, namely, CEDAR and NIST, and obtained 99.09 percent and 99.54 percent correct recognition rates al no-rejection level, respectively. The latter result was obtained by testing on more than 170,000 numerals.

引用

页码：535 / 539

页数：5

共 18 条

[1]

Baird H.S., 1992, Structured Document Image Analysis, P546

[2]

Chow CK., 1957, IRE T ELECT COMPUT E, VEC-6, P247, DOI DOI 10.1109/TEC.1957.5222035

[3]

Drucker H., 1993, International Journal of Pattern Recognition and Artificial Intelligence, V7, P705, DOI 10.1142/S0218001493000352

[4]

Dudani S. A., 1976, IEEE Transactions on Systems, Man and Cybernetics, VSMC-6, P325, DOI 10.1109/TSMC.1976.5408784

[5]

GUO J, 1993, IEICE D, V176, P835

[6]

HA TM, 1994, P 4 INT WORKSH FRONT, P97

[7]

HA TM, 1996, IAM96014 U BERN

[8] A DATABASE FOR HANDWRITTEN TEXT RECOGNITION RESEARCH [J].

HULL, JJ .

IEEE TRANSACTIONS ON PATTERN ANALYSIS AND MACHINE INTELLIGENCE, 1994, 16 (05) :550-554

[9]

KNERR S, 1993, P 1993 INT S NONL TH, V3

[10]

Lee D.S., 1993, Proceedings of the Third International Workshop on Frontiers in Handwriting Recognition, P153

← 1 2 →