Recognition of line segments with unevenness used in OCR and fingerprints

被引：10

作者：

Bourbakis, NG ^{[1
]}

Goldman, D

机构：

[1] SUNY Binghamton, Ctr Intelligent Syst, Binghamton, NY 13902 USA

[2] Tech Univ Crete, Dept ECE, Chania 73100, Crete, Greece

来源：

ENGINEERING APPLICATIONS OF ARTIFICIAL INTELLIGENCE | 1999年 / 12卷 / 03期

关键词：

line segments recognition with unevenness; OCR; handwriting recognition; fingerprints and recognition;

D O I：

10.1016/S0952-1976(98)00063-3

中图分类号：

TP [自动化技术、计算机技术];

学科分类号：

0812 ;

摘要：

This paper deals with the recognition of digital line segments with unevenness, and their application to the recognition of handwritten characters and fingerprints. The recognition methodology for line segments with unevenness is based on a heuristic criterion, where the user hierarchically defines the levels of tolerance (or acceptance) whereby an unevenness is removed from a line segment (ignored) or not. The results of this line-recognition approach are used on a graph-based method for the recognition of handwritten characters, where the graph nodes are straight or curve line segments. In addition, the same approach is used on the recognition of line segments extracted from fingerprints to form the appropriate graphs, which represent a fingerprint. (C) 1999 Published by Elsevier Science Ltd. All rights reserved.

引用

页码：273 / 279

页数：7