Handwritten Farsi (Arabic) word recognition: a holistic approach using discrete HMM

被引：101

作者：

Dehghan, M

Faez, K

Ahmadi, M ^{[1
]}

Shridhar, M

机构：

[1] Univ Windsor, Dept Elect & Comp Engn, Windsor, ON N9B 3P4, Canada

[2] Amirkabir Univ Technol, Dept Elect Engn, Tehran, Iran

[3] Univ Michigan, Dept Elect & Comp Engn, Dearborn, MI 48128 USA

来源：

PATTERN RECOGNITION | 2001年 / 34卷 / 05期

关键词：

handwritten word recognition; hidden Markov model; self-organizing feature map; parameter smoothing; Farsi/Arabic handwriting recognition;

D O I：

10.1016/S0031-3203(00)00051-0

中图分类号：

TP18 [人工智能理论];

学科分类号：

081104 ; 0812 ; 0835 ; 1405 ;

摘要：

A holistic system for the recognition of handwritten Farsi/Arabic words using right-left discrete hidden Markov models (HMM) and Kohonen self-organizing vector quantizatiun is presented. The histogram of chain-code directions of the image strips, scanned from right to left by a sliding window, is used as feature vectors. The neighborhood information preserved in the self-organizing feature map (SOFM), is used for smoothing the observation probability distributions of trained HMMs. Experiments carried out on test samples show promising performance results. (C) 2001 Pattern Recognition Society. Published by Elsevier Science Ltd. All rights reserved.

引用

页码：1057 / 1065

页数：7

共 11 条

[1] SURVEY AND BIBLIOGRAPHY OF ARABIC OPTICAL TEXT RECOGNITION [J].