Syntactic methodology of pruning large lexicons in cursive script recognition

被引：18

作者：

Madhvanath, S

Krpasundar, V

Govindaraju, V ^{[1
]}

机构：

[1] SUNY Buffalo, Dept Comp Sci, CEDAR, Amherst, NY 14228 USA

[2] Viewlog Syst Inc, W Marlboro, MA 01752 USA

[3] IBM, Almaden Res Ctr, San Jose, CA 95120 USA

来源：

PATTERN RECOGNITION | 2001年 / 34卷 / 01期

关键词：

off-line handwritten word recognition; lexicon reduction; holistic matching; shape descriptor; trie organization; elastic matching; postprocessing; cursive script;

D O I：

10.1016/S0031-3203(99)00201-0

中图分类号：

TP18 [人工智能理论];

学科分类号：

081104 ; 0812 ; 0835 ; 1405 ;

摘要：

In this paper, we present a holistic technique for pruning of large lexicons for recognition of off-line cursive script words. The technique involves extraction and representation of downward pen-strokes from the off-line cursive word to obtain a descriptor which provides a coarse characterization of word shape. Elastic matching is used to match the image descriptor with "ideal" descriptors corresponding to lexicon entries organized as a trie of stroke classes. On a set of 23,335 real cursive word images the reduction is about 70% with accuracy above 75%. (C) 2000 Pattern Recognition Society. Published by Elsevier Science Ltd. All rights reserved.

引用

页码：37 / 46

页数：10