DESIGN OF A NEURAL NETWORK CHARACTER RECOGNIZER FOR A TOUCH TERMINAL

被引：72

作者：

GUYON, I ^{[1
]}

ALBRECHT, P ^{[1
]}

LECUN, Y ^{[1
]}

DENKER, J ^{[1
]}

HUBBARD, W ^{[1
]}

机构：

[1] AT&T BELL LABS,HOLMDEL,NJ 07733

来源：

PATTERN RECOGNITION | 1991年 / 24卷 / 02期

关键词：

CHARACTER RECOGNITION; ONLINE CHARACTER RECOGNITION; HANDWRITTEN CHARACTERS; NEURAL NETWORKS; TOUCH TERMINAL; TOUCH SCREEN;

D O I：

10.1016/0031-3203(91)90081-F

中图分类号：

TP18 [人工智能理论];

学科分类号：

081104 ; 0812 ; 0835 ; 1405 ;

摘要：

We describe a system which can recognize digits and uppercase letters handprinted on a touch terminal. A character is input as a sequence of [x(t), y(t)] coordinates, subjected to very simple preprocessing, and then classified by a trainable neural network. The classifier is analogous to "time delay neural networks" previously applied to speech recognition. The network was trained on a set of 12,000 digits and uppercase letters, from approximately 250 different writers, and tested on 2500 such characters from other writers. Classification accuracy exceeded 96% on the test examples.

引用

页码：105 / 119

页数：15

共 14 条

[1]

[Anonymous], 1987, LEARNING INTERNAL RE

[2] What Size Net Gives Valid Generalization? [J].

Baum, Eric B. ;

Haussler, David .

NEURAL COMPUTATION, 1989, 1 (01) :151-160

[3]

BECKER S, 1988, CRGTR885 U TOR CONN

[4]

BOTTOU LY, 1988, THESIS U PARIS 5

[5]

Denker J., 1987, Complex Systems, V1, P877

[6]

LANG K, 1988, CMUCS88152 CARN U TE

[7]

LECUN Y, 1989, IEEE COMMUN MAG NOV, P41

[8]

LeCun Y., 1989, CONNECTIONISM PERSPE

[9] METHODOLOGIES IN PATTERN-RECOGNITION AND IMAGE-ANALYSIS - A BRIEF SURVEY [J].

MANTAS, J .

PATTERN RECOGNITION, 1987, 20 (01) :1-6

[10]

Nagy G, 1982, HDB STATISTICS, V2, P621

← 1 2 →