Handwritten Character Recognition : A Comprehensive Review on Geometrical Analysis
Handwritten Character Recognition : A Comprehensive Review on Geometrical Analysis
复制标题
手写字符识别:几何分析综合评述
DOI：
--
复制
发表时间：
2015
期刊：
影响因子：
--
通讯作者：
Meenu Mohan
中科院分区：
文献类型：
--
作者：
Meenu Mohan
本文详细回顾了离线手写字符识别。 HCR 是一种光学字符识别，它将人类可读的字符转换为机器可读的格式。在 HCR 中，要达到 99% 的准确率是非常困难的。这里对字符识别中特征提取的几何方法进行了详细研究，重点关注基于区域的技术，并分析了 HCR 的效率取决于选择适当的特征提取方法和分类器。对字符识别中的预处理、分割、特征提取和分类等各个步骤进行了比较研究。还分析了 HCR 的各种应用领域，如邮政地址读取、邮件分拣、文本输入办公自动化、人员识别、签名验证、银行支票处理等。
This paper presents a detailed review of Offline Handwritten Character Recognition. HCR is an optical character recognition, which convert the human readable character to machine readable format. In HCR, to attain 99% accuracy is very difficult. Here a detailed study on Geometrical methods of feature extraction in character recognition has been done by giving more emphasis to Zone based techniques and it has been analyzed that the efficiency of HCR depends on the selection of appropriate feature extraction methods and classifier. A comparative study in various steps in character recognition like Preprocessing, Segmentation, Feature Extraction and Classification are carried out. Various application areas of HCR like Postal address reading, mail sorting, office automation for text entry, person identification, signature verification, bank-check processing etc. are also analyzed.