改进的Stein无偏风险估计的小波滤波算法  

The improved wavelet filtering algorithm based on Stein’s unbiased risk estimation

作　　者：刘先 边少锋 翟国君 李邵波 LIU Xian;BIAN Shaofeng;ZHAI Guojun;LI Shaobo(Key Laboratory of Geophysical Survey and Evaluation of Ministry of Education,China University of Geosciences,Wuhan 430074,China;School of Geography and Information Engineering,China University of Geosciences,Wuhan 430074,China)

机构地区：[1]中国地质大学(武汉)地质探测与评估教育部重点实验室,武汉430074 [2]中国地质大学(武汉)地理与信息工程学院,武汉430074

出　　处：《测绘科学》2024年第12期158-166,共9页Science of Surveying and Mapping

基　　金：国家自然科学基金项目(42374050,42342024);地质探测与评估教育部重点实验室主任基金项目(GLAB2024ZR05);中央高校基本科研业务费项目。

摘　　要：针对常用的阈值法无法适应短信号的去噪,且传统Stein无偏风险估计自适应阈值法存在理论缺陷的问题,该文改进了基于Stein无偏风险估计的自适应阈值法,并将原始小波系数与纯净小波系数的均方差估计推广到一般形式。通过仿真实验对比其他阈值法的指标发现改进算法对短信号有更稳定的去噪效果,其他场景下效果与其他方法相当。该文方法提高了数据缺失或波形较短情况下的波形去噪精度,较好地适应于极浅水测深数据,为该场景下的测深数据处理提供了新思路。Noting that common threshold methods fail to effectively denoise short signals,and the traditional Stein's Unbiased Risk Estimation adaptive threshold method has theoretical deficiencies,the improved wavelet filtering algorithm based on Stein's unbiased risk estimation was proposed in this paper,and extending the mean square error estimation of both the original and pure wavelet coefficients to a general form.Through simulation experiments comparing the indicators of other threshold methods,it was found that improved algorithm has a more stable denoising effect on short length data,and the denoising effect in other scenarios is similar to other methods.Improved the accuracy of waveform denoising algorithms in cases of missing data or short waveforms,making it well suited for shallow water depth measurement data.

关 键 词：机载激光测深 小波阈值去噪 Stein无偏风险估计 高斯函数拟合 

分 类 号：P229[天文地球—大地测量学与测量工程]