An improved threshold function is presented based on critical region and adjustable parameters of shrinkage.While some of the threshold functions which have been brought forward by others, don't have high level of differentiability and shrinkage flexibility. The critical region, in which the wavelet coefficients are processed meticulously, is defined by the probability density distribution of white noise and the estimation of threshold. And the adjustable parameters of shrinkage, which determine the contraction degree of wavelet coefficients, are set up to improve the efficiency of filter. Particularly, according to the experiment, the new function can keep a good balance between noise elimination and edge feature holding, and the better de-noising effect is achieved.%针对现有的小波阈值函数在阈值邻域内不具有高阶可导性,小波系数收缩缺乏灵活性,在此构造了一种具有过渡临界区域和收缩调和参数的阈值函数.根据白噪声的概率密度分布和滤波阈值的估计值确定过渡临界区域,细化阈值邻城内小波系数的处理.通过收缩调和参数调整小波系数的收缩程度,以提高滤波信号的信噪比.实验结果表明,在此构造的小波阈值函数能够在消噪和保留原有信号的弱特征之间获得较好的平衡,从而改善小波阈值滤波算法的性能.
展开▼
