摘要
本文提出了在有相当强烈的干扰条件下,基于Haar小波变换的印刷体汉字处理的一种新方法。此方法只对反映原始字符信号概貌的小波变换低频部分保留,同时忽略包含大量干扰信号的高频部分,由此达到剔除大部分干扰信号的效果。实验结果表明本文提出的方法具有良好的去干扰特性,为汉字处理提供了新的思路。
In this paper,a novel approach based on Haar wavelet transform is proposed to process typography Chinese characters which are intensively disturbed, This approach addresses eliminating mostly of signals disturbing characters by only reserving low-frequency portion of result of wavelet transform which carries general picture of original character, at the same time, ignoring high-frequency portion of result of wavelet transform which carries mostly of disturbing signals. The result of experiment shows that the approach proposed by this paper has the feature eliminating intensively-disturbing easily and effectively, and it provides Chinese character processing with new idea.
出处
《计算机应用与软件》
CSCD
北大核心
2005年第8期46-47,111,共3页
Computer Applications and Software
