维吾尔文脱机手写识别建模研究  

Research of Offline Handwriting Recognition's Modeling in Uighur

作　　者：黄弋石[1] 

机构地区：[1]南通大学公共卫生学院实验中心,江苏南通226019

出　　处：《信息记录材料》2018年第2期89-91,共3页Information Recording Materials

摘　　要：为了解决维吾尔文的印刷文字与手写文字的脱机识别,本文提出了新颖识别技术。规定了核心模型定义,使用五种机器读法,来识别文字的特征。使用简捷方便的编码方法,使得模型结构舒畅美观。对于常用文字的编码,没有重码。在理论上,实现了眼睛能够识别则机器即可辨认的目标。显然,这种模型设计,也可以改善其他研究者的方案,大幅度提高对应效率与成功率。In order to solve the Uighur printed text with handwritten words offline recognition, this paper proposes new recognition technology. The core model definition is defined, and five machine pronunciations are used to identify the features of the text. Using simple and convenient coding method, the model structure is pleasant and beautiful. For text coding, there is no duplicate code. In theory, the target is recognized by the eyes and the machine can recognize it. Obviously, this model design can also improve the other researchers＇ program, and greatly improve the corresponding efficiency and success rate.

关 键 词：维吾尔文手写识别 手写识别 脱机识别 连笔识别 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]