基于改进阈值的地质雷达图象多小波去噪

Image De-noising of Ground Penetrating Radar in Multiwavelet Transform Based on Improved Thresholding Function

探讨了基于DG H M多小波、多分辨分析下的地质雷达(GPR)图象阈值化去噪方法,在分析Donoho D L和Johnstone I M提出的小波阈值去噪方法的基础上,提出了一个改进的阈值函数,并对不同阈值函数下的去噪效果进行了比较分析。实验结果表明,利用这一改进的阈值函数和DGH M多小波对GPR图象进行去噪,可获得比传统软、硬阈值方法更好的效果。

The algorithm for noise reduction of Ground Penetrating Radar(GPR)image with DGHM multiwavelet Multiple Resolution Analysis(MRA)is discussed in this paper.Based on analysis of the wavelet threshold denoising which is put forward by Donoho D L and Johnstone I M,an improved thresholding function is presented.The de-noising of GPR image based on different thresholding function is analyzed and compared.The results of simulation experiment indicate that the de-noising effect of GPR image with multiwavelet based on this new thresholding function is better than DJ' soft- and hard- thresholding methods.