摘要
在模式识别中 ,特征向量的选择与提取至关重要 .本文讨论用统计方法进行文字识别的有关问题 ,利用小波分解及分形维的概念 ,建立了文字识别的特征向量的提取方法 ,这种方法将文字的二维图形化为一维图形 ,对得到的一维图形 (曲线 )进行小波分解 ,计算少数几个分解得到的曲线的分形维数 ,以它们构成特征向量 .并对有关文字做了试验 。
In the pattern recognition,it is important to select and extract proper feature vector.In the paper,it is discussed that associated contents of literal recognition using the statistical classification method.Using wavelet decomposition and fractal dimension,the feature vector of literal is extracted.This method transforms 2_D graphic literals into 1_D graph which is decomposed by wavelet to get curves,then the feature vector is formed by calculating fractal dimension of several segments of curves.Some charcaters were tested using the method,the result is satisfied.
出处
《西安工业学院学报》
2004年第3期226-230,239,共6页
Journal of Xi'an Institute of Technology
关键词
文字识别
小波分解
特征向量
分形维
literal recognition
wavelet decomposition
feature vector
fractal dimension
