基于连续小波变换去除奇异信号方法

Removing singular signals based on continuous wavelet transforms

提出了高维小波变换的一般形式,在一系列的多维小渡变换中,选择一个适当的小波变换,对小波正变换和逆变换进行离散,构造简单易行且有效的数值计算方法．仿真实例表明:对同时带有白噪声和野值的二维信号进行离散的小波正变换和小波逆变换处理,可以彻底地去除信号中白噪声和野值,重建原有信号．

The general form of multidimensional wavelet transform was put forward. Among a series of multidimensional wavelet transforms, one proper wavelet transform was selected. By discretizing the transform of wavelet and the inverse transform of wavelet, a simply and effective numeric arithmetic can be constructed. Simulation instances showed, by dealing the 2 - dimension signals including the white noises and Outlier at the same time with discretized the transform of wavelet and the inverse transform of wavelet, the white noises and Outlier in the signals can be completely wiped off, and the original signals was reconstructed.