摘要
为了克服硬阈值函数不连续,软阈值函数中估计小波系数与分解小波系数之间存在着恒定偏差的缺陷,更好的改进滤波效果,提高去噪质量,通过对硬软阈值去噪方法的分析,比较了硬软阈值去噪的方法,构造了一个新的阈值函数。新阈值函数表达式简单易于计算,克服了硬软阈值的缺点,而且是高阶可导的,便于进行各种数学处理。仿真结果表明:无论是视觉效果,还是信噪比增益和最小均方差均优于传统的硬软阈值方法。仿真试验证实了该改进方案的有效性和优越性。`这种方法可以广泛应用于电力系统、语音信号去噪等领域。
In order to overcome the discontinuance of the hard thresholding fanction and the constant deviation between the estimated wavelet coefficients and the decomposition wavelet coefficients in the soft thresholding function, and improve the filtering effect and the quality of denoising , after analysing the methods of the hard and soft threshold denoising and comparing the methods of hard and soft threshold denoising, a new threshold function is construeted. New threshold function expression is simple and easy to calculate and overcomes the shortcomings of the hard and soft thresholding function. It is also a higher - order derivative and facilitates various mathematical treatment. Simulation results show that：both the visual effects and the SNR gains and minimum mean square error are superior to the traditional method of the soft and hard threshold. Simulation experiments confirm that the programme is effective and superior. This method can be widely used in power systems and voice signal denoising fields.
出处
《计算机仿真》
CSCD
北大核心
2009年第4期348-351,共4页
Computer Simulation
基金
江苏省镇江市科学技术局基金资助项目(BG2007033)
关键词
小波变换
小波阈值去噪
阈值函数
均方误差
信噪比
Wavelet transform
Wavelet denoising threshold
Threshold function
Mean square error
SNR
