RECOGNITION OF HANDWRITTEN AND MACHINE-PRINTED TEXT FOR POSTAL ADDRESS INTERPRETATION

被引：64

作者：

SRIHARI, SN

机构：

[1] Center of Excellence for Document Analysis and Recognition (CEDAR), State University of New York at Buffalo, Buffalo

来源：

PATTERN RECOGNITION LETTERS | 1993年 / 14卷 / 04期

关键词：

OCR; ADDRESS INTERPRETATION; HANDWRITING RECOGNITION; NEURAL NETWORKS;

D O I：

10.1016/0167-8655(93)90095-U

中图分类号：

TP18 [人工智能理论];

学科分类号：

081104 ; 0812 ; 0835 ; 1405 ;

摘要：

Postal address interpretation is the task of assigning to letter mail pieces a delivery point encoding. e.g., ZIP+4 Code. The encoding is determined from images of destination addresses on mail piece faces: addresses that are handwritten, are of poor-quality machine printing, are incomplete or incorrect. This paper describes several recognition algorithms used in the interpretation of handwritten and machine-printed address text (digits/symbols/alphabets/words).

引用

页码：291 / 302

页数：12

共 16 条

[1]

Cohen E., 1991, International Journal of Pattern Recognition and Artificial Intelligence, V5, P221, DOI 10.1142/S0218001491000156

[2]

CULLEN PB, 1992, EIGHTH CONFERENCE ON ARTIFICIAL INTELLIGENCE FOR APPLICATIONS : PROCEEDINGS, P127

[3]

D'Amato D., 1982, Proceedings of PRIP 82. IEEE Computer Society Conference on Pattern Recognition and Image Processing, P165

[4] A COMBINATION OF STATISTICAL AND SYNTACTICAL PATTERN-RECOGNITION APPLIED TO CLASSIFICATION OF UNCONSTRAINED HANDWRITTEN NUMERALS [J].

DUERR, B ;

HAETTICH, W ;

TROPF, H ;

WINKLER, G .

PATTERN RECOGNITION, 1980, 12 (03) :189-199

[5]

FENRICH R, 1992, PIXELS FEATURES, V3, P47

[6] HOW NEURAL NETWORKS LEARN FROM EXPERIENCE [J].

HINTON, GE .

SCIENTIFIC AMERICAN, 1992, 267 (03) :145-151

[7]

HO TK, 1992, THESIS SUNY BUFFALO

[8]

KUAN CC, 1991, Patent No. 5058182

[9]

LEE DS, 1991, ARTIFICIAL NEURAL NE, P89

[10]

PALUMBO PW, 1992, IEEE COMPUT, P34

← 1 2 →