基于局部保持投影与隐马尔可夫模型的维文字符识别

doi:10.3724/SP.J.1087.2012.02309

计算机应用 ›› 2012, Vol. 32 ›› Issue (08): 2309-2312.DOI: 10.3724/SP.J.1087.2012.02309

基于局部保持投影与隐马尔可夫模型的维文字符识别

刘卫¹,李和成²

1. 青海师范大学物理系，西宁 810008
2. 青海师范大学数学系，西宁 810008

收稿日期:2012-01-16 修回日期:2012-03-22 发布日期:2012-08-28 出版日期:2012-08-01
通讯作者: 刘卫
作者简介:刘卫(1975-)，男，山东滕州人，副教授，主要研究方向：图像处理、模式识别、机器学习;
李和成(1973-)，男，青海乐都人，教授，博士，主要研究方向：进化计算、数据挖掘。
基金资助:
复杂双层规划问题的高性能可信进化算法研究

Uighur characters recognition based on locality preserving projection and hidden Markov model

LIU Wei¹,LI He-cheng²

1. Department of Physics, Qinghai Normal University, Xining Qinghai 810008, China
2. Department of Mathematics, Qinghai Normal University, Xining Qinghai 810008, China

Received:2012-01-16 Revised:2012-03-22 Online:2012-08-28 Published:2012-08-01
Contact: LIU Wei

摘要/Abstract

摘要： 针对传统隐马尔可夫模型(HMM)在对手写维吾尔文字符建模时，字符宽度变化大，模型训练收敛缓慢，且易陷入局部极值的问题，提出一种基于保局投影(LPP)与HMM相结合的维吾尔字符识别方法。首先，通过高度归一化保持原图像的宽高比，用滑动窗获取子图像序列，形成观测向量序列；其次，采用局部保持投影将观测序列映射到低维空间，并用随机抽样方法降低邻接图矩阵的规模；最后，采用新观测序列训练HMM。该算法在降维的同时提高了HMM的收敛速度，降低了陷入局部极值的风险。实验结果显示，算法的平均收敛步数减少，错误率降低，表明算法是有效的。

关键词: 隐马尔可夫模型, 局部保持投影, 维文识别, 归一化, 收敛

Abstract: Concerning the shortcomings of classical Hidden Markov Model (HMM) in handwritten Uighur characters recognition, such as largly varied width of characters, slow convergent speed and premature convergence, a new Uighur characters recognition algorithm was proposed in combination with Locality Preserving Projection (LPP) and HMM. Firstly, the aspect ratio of original image was maintained by a highly-normalized method. Sub-images were obtained by using sliding window, and observation sequences were extracted from these windows. Secondly, the observation sequences were mapped into low-dimensional space based on LPP, and the scale of adjacency matrix was reduced via the random sampling technique. Finally, HMM was trained by adopting obtained observation sequences. The algorithm decreases dimension of observation vectors, accelerates the convergence, and prevents premature convergence effectively. The simulation results show the LPP-HMM algorithm is efficient and robust, which decrease average convergence steps as well as errors.

Key words: Hidden Markov Model (HMM), Locality Preserving Projection (LPP), Uighur characters recognition, normalization, convergence

中图分类号:

TP391.41

刘卫李和成. 基于局部保持投影与隐马尔可夫模型的维文字符识别[J]. 计算机应用, 2012, 32(08): 2309-2312.

LIU Wei LI He-cheng. Uighur characters recognition based on locality preserving projection and hidden Markov model[J]. Journal of Computer Applications, 2012, 32(08): 2309-2312.

参考文献

[1]赵继印，郑蕊蕊，吴宝春,等. 脱机手写体汉字识别综述[J].电子学报, 2010, 38(2): 405-414. [2]袁保社，吾守尔·斯拉木. 一种手写维吾尔文字母识别算法[J].计算机工程,2010, 36(2): 186-188. [3]UBUL K, HAMDULL A, AYSA A, et al. Research on Uyghur off-line handwriting-based writer identification [C]// ICSP 2008: 9th International Conference on Signal Processing. Piscataway: IEEE, 2008: 1656-1659. [4]哈力木拉提，阿孜古丽. 多字体印刷维吾尔文字符识别系统的研究与开发[J].计算机学报,2004, 27(11): 1480-1484. [5]RABINER L R. A tutorial on hidden Markov models and selected applications in speech recognition[J].Proceedings of the IEEE， 1989，77(2): 257-286. [6]LORIGO L M, GOVINDARAJU V. Off-line Arabic handwriting recognition: a survey[J].IEEE Transactions on Pattern Analysis and Machine Intelligence,2006, 28(5): 712-724. [7]VINCIARELLI A. A survey on off-line cursive word recognition[J].Pattern Recognition, 2002, 35(7): 1433-1446. [8]XIANG DONG，YAN HUAHUA，CHEN XIANQIAO, et al. Offline Arabic handwriting recognition system based on HMM [C]// 2010 3rd IEEE International Conference on Computer Science and Information Technology. Piscataway: IEEE, 2010, 1: 526-529. [9]MOHAMAD R A-H, LIKFORMAN-SULEM L, MOKBEL C. Combining slanted-frame classifiers for improved HMM-based Arabic handwriting recognition[J].IEEE Transactions on Pattern Analysis and Machine Intelligence, 2009, 31(7): 1165-1177. [10]EL-HAJJ R, LIKFORMAN-SULEM L, MOKBEL C. Arabic handwriting recognition using baseline dependant features and hidden Markov modeling [C]// Proceedings of the Eighth International Conference on Document Analysis and Recognition. Piscataway: IEEE, 2005, 2: 893-897. [11]VINCIARELLI A, BENGIO S. Offline cursive word recognition using continuous density hidden Markov models trained with PCA or ICA features [C]// ICPR'02: Proceedings of 16th International Conference on Pattern Recognition. Washington, DC: IEEE Computer Society, 2002, 3: 81-84. [12]HE XIAOFEI, NIYOGI P. Locality preserving projection [C]// Advances in Neural Information Processing Systems. Cambridge: MIT Press, 2004: 153-160. [13]黄勇. 基于图像优化局部保留投影的人脸表情识别[J].计算机工程与应用, 2011, 47(27): 210-211. [14]刘敏，李晓东，王振海. 一种新的有监督保局投影人脸识别算法[J].JOCA, 2009, 29(5): 1416-1418.

[1]	甘岚, 沈鸿飞, 王瑶, 张跃进. 基于改进DCGAN的数据增强方法[J]. 计算机应用, 2021, 41(5): 1305-1313.
[2]	蒋考林, 白玮, 张磊, 陈军, 潘志松, 郭世泽. 基于多通道图像深度学习的恶意代码检测[J]. 计算机应用, 2021, 41(4): 1142-1147.
[3]	魏博, 杨茸, 舒思豪, 万勇, 苗建国. 基于离子运动-人工蜂群算法的移动机器人路径规划[J]. 计算机应用, 2021, 41(2): 379-383.
[4]	乔钢柱, 王瑞, 孙超利. 基于分解的高维多目标改进进化算法[J]. 《计算机应用》唯一官方网站, 2021, 41(11): 3097-3103.
[5]	张驰, 李铸洪, 刘舟, 沈未名. 基于场景图划分的无人机影像定位算法[J]. 计算机应用, 2021, 41(10): 3004-3009.
[6]	李二超, 李康伟. 基于最小距离和聚合策略的分解多目标进化算法[J]. 计算机应用, 2021, 41(1): 22-28.
[7]	李二超, 杨蓉蓉. 基于自适应反向学习的多目标分布估计算法[J]. 计算机应用, 2021, 41(1): 15-21.
[8]	霍纬纲, 王慧芳. 基于自编码器和隐马尔可夫模型的时间序列异常检测方法[J]. 计算机应用, 2020, 40(5): 1329-1334.
[9]	刘静姝, 王莉, 刘惊雷. 无需特征分解的快速谱聚类算法[J]. 计算机应用, 2020, 40(12): 3413-3422.
[10]	娄梦莹, 王天景, 刘娅琴, 杨丰, 黄靖. 基于侧链连接卷积神经网络的手掌静脉图像识别[J]. 计算机应用, 2020, 40(12): 3673-3678.
[11]	陈维兴, 刘清涛, 孙习习, 陈斌. 机坪感知网络的快速收敛平均一致性时间同步算法[J]. 计算机应用, 2020, 40(11): 3407-3412.
[12]	高媛, 王晓晨, 秦品乐, 王丽芳. 基于深度可分离卷积和宽残差网络的医学影像超分辨率重建[J]. 计算机应用, 2019, 39(9): 2731-2737.
[13]	马方方, 刘树波, 熊星星, 牛晓光. 可穿戴设备数值型敏感数据本地差分隐私保护[J]. 计算机应用, 2019, 39(7): 1985-1990.
[14]	杨宏宇, 李博超. 基于逆向习得推理的网络异常行为检测模型[J]. 计算机应用, 2019, 39(7): 1967-1972.
[15]	韦姗姗, 胡圣波, 鄢婷婷, 莫金容. 基于分步小波变换的对流层电波传播特性和分析软件[J]. 计算机应用, 2019, 39(6): 1792-1798.

基于局部保持投影与隐马尔可夫模型的维文字符识别

Uighur characters recognition based on locality preserving projection and hidden Markov model

PDF

可视化

摘要/Abstract

引用本文

使用本文

参考文献

相关文章 15

编辑推荐

Metrics