摘要
针对硬阈值去噪效果不佳,软阈值过度光滑使信号失真的缺点,本文在研究Donoho的软、硬阈值去噪方法基础上,提出了一种新的阈值函数,该阈值函数在整个定义域内统一定义,表达式简单易于计算,同软阈值函数一样具有连续性。通过对典型信号Blocks、Bumps、Heavy sine、Doppler的仿真实验结果显示,与基于硬、软阈值去噪方法相比,基于新阈值函数的小波去噪方法取得非常好的去噪效果,在均方误差等性能指标上表现优异,重建信号的信噪比也得到提高。
Aiming at the disadvantages that hard threshold is poor in de-noising and soft threshold makes signal distortion because of smooth transition, this paper, based on the soft and hard thresholding method put forward by Donoho, proposes a new thresholding function. The threshold function unifies definition in the whole domain,making the expression easy to calculate, and as continuous as the soft-thresholding function. The results of the simulation experiment for the typical signal of Blocks, Bumps, Heavy sine, and Doppler show that the wavelet de-noising method of new threshold function has a good de-noising effect, and does well on the performance index of mean-square error. The signal to noise ratio of signal reconstruction is also improved.
出处
《科技广场》
2015年第9期20-23,共4页
Science Mosaic
关键词
小波变换
阈值函数
去噪
小波阈值去噪
信噪比
均方误差
Wavelet Transform
Threshold Function
De-Nosing
Wavelet Threshold De-Noising
Signal To noise Ratio(SNR)
Mean Square Error(MSE)