基于提升小波的改进半软阈值降噪方法

Improved Half-soft Threshold Denoising Based on Lifting Wavelet

针对小波变换下的传统阈值降噪方法的缺点,提出了一种基于提升小波变换的改进半软阈值方法,通过提升小波变换对含噪声信号进行分解降噪,并利用改进的半软阈值方法对分解过程中的高频信号进行阈值降噪处理。对提升小波变换下的传统阈值和改进半软阈值方法的仿真试验进行比较,证明了后者在信号降噪时,相比传统的阈值方法能够获得更高的信噪比增益,且运算速度较快,更适合于实际应用。

When the lifting wavelet is used to reduce noises in different signals, thresholds of wavelet coefficient must be determined to drop the small coefficient. In order to overcome shortcomings of thresholds used in the classical wavelet, improved half-soft threshold based on the lifting wavelet is proposed for signal denoising in this paper. According to the simulation results based on the lifting wavelet, a higher signal to noise ratio （SNR） gain and a faster processing are achieved by using improved half-soft threshold, compared with those achieved by using classical thresholds.