一种基于改进阈值函数的小波阈值降噪算法  

A new wavelet threshold de-noising method based on improved threshold function

作　　者：倪培峰[1] 胡雄[1] 

机构地区：[1]上海海事大学物流工程学院,上海201306

出　　处：《电子技术应用》2016年第8期98-100,104,共4页Application of Electronic Technique

基　　金：上海海事大学研究生创新基金资助项目(YXR2015122)

摘　　要：针对小波阈值降噪中硬阈值函数和软阈值函数的不足,结合现有文献提出一种新的阈值函数。新阈值函数克服了传统阈值函数的缺点,保证了阈值函数的连续性,同时可以通过改变参数灵活地调节函数。在新阈值函数的基础上结合改进的阈值确定方法,提出一种新的降噪算法。通过MATLAB仿真,对几种小波降噪算法进行了试验分析,利用信噪比和均方根误差两个指标进行评价。结果表明,相比于传统的降噪算法,新降噪算法取得了更好的降噪效果。Aiming at the defects of hard threshold function and soft threshold function in wavelet threshold de- noising algorithm, a new threshold function was proposed combining with the existing literatures. The new threshold function can overcome the weakness of traditional threshold functions, and guarantee the continuity of threshold function. At the same time, the function can be flexibly adjusted by changing the parameters. Based on the new threshold function, a new wavelet threshold de- noising algorithm was pro-posed combing with an improved threshold selection method. Though MATLAB simulation, several kinds of wavelet de- noising algo-rithm were tested and analyzed, using signal- to- noise ratio and root- mean- square- error as two indicators for evaluation. The re-sults showed that the new noise reduction algorithm proposed obtained better noise reduction effect compared to the traditional de-noising algorithm.

关 键 词：小波降噪 阈值函数 阈值选取 信噪比 均方根误差 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]