Abstract: Based on the multi-analysis wavelet threshold denoising method which is put forward by D.L.Dohono and I.M.Johnstone，this paper puts forward a new threshold function.The new threshold function has many advantages over the classical soft or hard threshold function.It is clear in physics meaning and simple in expression.The results show that this improved method is effective in removing white noise，gives better RMSE performance and SNR gains than the soft or hard threshold methods，makes it possible to embody the advantages of the wavelet threshold denoising method more fully．

Key words: wavelet transform, hard-threshold denoising, soft-threshold denoising, threshold function

摘要： 在D.L.Dohono和I.M.Johnstone提出的多分辨分析小波阈值去噪方法的基础上，提出了一种新的阈值函数。与传统的硬阈值和软阈值比，该函数不仅易于计算，而且具有优越的数学特性和清晰的物理意义。实验结果表明，该方法可以有效地去除白噪声干扰，无论在视觉效果上还是在信噪比和均方误差定量指标上均明显优于常用的软、硬阈值及改进的软硬阈值折中算法，充分体现出小波阈值去噪方法的优越性。

关键词: 小波变换, 硬阈值去噪, 软阈值去噪, 阈值函数