Off-line Arabic handwritten characters recognition based on a Hidden Markov Models

被引：0

作者：

Amrouch, M. ^{[1
]}

Elyassa, M. ^{[1
]}

Rachidi, A. ^{[1
]}

Mammass, D. ^{[1
]}

机构：

[1] IRF SIC Lab, Fac Sci, Agardir, Morocco

来源：

IMAGE AND SIGNAL PROCESSING | 2008年 / 5099卷

关键词：

isolated Arabic hand-written characters; Hough transformation; Hidden Markov Models;

D O I：

暂无

中图分类号：

TP18 [人工智能理论];

学科分类号：

081104 ; 0812 ; 0835 ; 1405 ;

摘要：

We present a system based on Hidden Markov Models (HMMs) for off-line isolated Arabic handwritten characters recognition. The Hough accumulator of the character image is partitioned into equal horizontal bands that will be used to extract directional information. This information is translated into sequences of observations that are used to train the model for each character during the learning step.

引用

页码：447 / 454

页数：8