MODEL-BASED RECOGNITION OF CHARACTERS IN TRADEMARK ARTWORK

被引：1

作者：

BROSSMAN, C

CROSS, GR

机构：

[1] Department of Computer Science, Washington State University, Pullman

来源：

PATTERN RECOGNITION LETTERS | 1990年 / 11卷 / 05期

关键词：

image databases; model-based recognition; multi-media databases; Multifont character recognition; similarity-based retrieval; thinning of digital patterns;

D O I：

10.1016/0167-8655(90)90046-5

中图分类号：

TP18 [人工智能理论];

学科分类号：

081104 ; 0812 ; 0835 ; 1405 ;

摘要：

We are building a legal expert system that contains a database of trademarks and reasons about their similarity to assess possible infringement. Our attention in this document is limited to the development of a character recognition component which determines the location and identity of alphabetic characters in the trademarks. The techniques include structural models, graph homeomorphisms, and thinning. Modifications and enhancements to these techniques were required along with changes to existing algorithms. © 1990.

引用

页码：363 / 370

页数：8

共 13 条

[1]

CAPLAN D, 1966, BRIT TRADEMARKS SYMB

[2]

CHARTLAND G, 1986, GRAPHS DIGRAPHS

[3]

CROSS GR, 1988, 4TH P INT C THEM LAW, P11

[4]

Fu K.S., 2019, APPL PATTERN RECOGNI

[5]

HIRATA T, 1980, LECTURE NOTES COMPUT, V108, P65

[6]

HOPCROFT JE, 1982, COMPLEXITY COMPUTER, P131

[7]

KWOK PCK, 1988, COMMUN ACM, V11, P1314

[8] STRUCTURAL DESCRIPTIONS AND INEXACT MATCHING [J].

SHAPIRO, LG ;

HARALICK, RM .

IEEE TRANSACTIONS ON PATTERN ANALYSIS AND MACHINE INTELLIGENCE, 1981, 3 (05) :504-519

[9]

SHAPIRO LG, 1980, IEE T PATTERN ANAL M, V1, P111

[10]

SHILLMAN RJ, 1977, P IEEE COMPUTER SOC, P327

← 1 2 →