首页 | 本学科首页   官方微博 | 高级检索  
     检索      

基于正交混合Gauss模型的脱机手写数字识别
引用本文:张睿,丁晓青,刘海龙.基于正交混合Gauss模型的脱机手写数字识别[J].清华大学学报(自然科学版),2002,42(1):19-22.
作者姓名:张睿  丁晓青  刘海龙
作者单位:清华大学,电子工程系,智能技术与系统国家重点实验室,北京,100084
基金项目:国家“八六三”高技术项目 ( 86 3-30 6 -ZT0 3-0 3-1),国家自然科学基金资助项目 ( 6 9972 0 2 4)
摘    要:在基于统计方法的脱机手写数字识别中 ,为更加有效地描述特征的类条件概率分布 ,设计出性能优良的 Bayes分类器 ,采用了混合 Gauss模型。为减少模型的参数 ,通常假设各 Gauss分量的协方差矩阵为对角阵。由于各维特征之间统计相关 ,因此需要大量的 Gauss分量才能较好地描述特征的类条件概率分布 ,使得混合模型的阶数较高。为降低模型的阶数 ,采用了正交混合 Gauss模型 ,即先对各类别的特征分别进行 KL 变换 ,再将变换后的特征用混合 Gauss模型来表示。其中混合 Gauss模型的参数可以通过 EM算法进行估计。最后 ,在 NIST (National Institute of Standards andTechnology)手写数字样本集上对该方法的识别性能进行了验证

关 键 词:混合Gauss模型  正交混合Gauss模型  脱机手写数字识别  字符识别
文章编号:1000-0054(2002)01-0019-04
修稿时间:2001年1月8日

Offline handwritten numeral recognition using the orthogonal Gaussian mixture model
ZHANG Rui,DING Xiaoqing,LIU Hailong.Offline handwritten numeral recognition using the orthogonal Gaussian mixture model[J].Journal of Tsinghua University(Science and Technology),2002,42(1):19-22.
Authors:ZHANG Rui  DING Xiaoqing  LIU Hailong
Abstract:The performance of the statistical approach to offline handwritten numeral recognition is improved using the Gaussian mixture model (GMM) to approximate an arbitrary class conditional probability density. For simplification, the GMM is commonly assumed to have diagonal covariance matrixes. Statistical corrrelation of the features of handwritten numerals requires a large number of mixture components to obtain a good approximation. To solve this problem, the feature vectors are first transformed to the space spanned by the eigenvectors of the covariance matrix to reduce the correlation among the elements. The GMM is then applied to the transformed feature vectors. This GMM is defined as the orthogonal Gaussian mixture model (OGMM) which gives a better approximation than GMM with the same number of mixture components. The OGMM parameters can be estimated by EM algorithm. The algorithm effectiveness is demonstrated by applying it to the NIST (National Institute of Standards and Technology) database.
Keywords:Gaussian  mixture model  orthogonal Gaussian mixture model  offline handwritten numeral recognition  character recognition
本文献已被 CNKI 万方数据 等数据库收录!
设为首页 | 免责声明 | 关于勤云 | 加入收藏

Copyright©北京勤云科技发展有限公司  京ICP备09084417号