小波阈值消噪方法分析及应用  

Analysis and Application of De-noising Method for Wavelet Thresholding

作　　者：沙磊[1,2] 任超[2] 彭家頔[2] 文鸿雁[2] 

机构地区：[1]安徽省地质实验研究所,合肥市芜湖路239号230001 [2]桂林理工大学土木与建筑工程学院,桂林市建干路12号541004

出　　处：《测绘信息与工程》2011年第6期15-17,共3页Journal of Geomatics

摘　　要：详细介绍了小波阈值消噪中阈值函数的选取和阈值的具体估计方法,并对阈值函数的选取进行了比较分析,以四川二滩大坝变形监测数据消噪前后的信噪比和均方根误差两个指标,对各种阈值方法进行了对比分析。理论分析、应用实例的结果表明,几乎硬阈值方法比Garrote阈值方法能更好地恢复被噪声湮没的局部变形信息。The estimation method and the function choosing in de-noising of wavelet threshold are introduced,and comparative analysis of threshold function choosing is presented.Based on observed data from Ertan Dam in Sichuan Province,comparison between original and estimated signal after de-noising on evaluate index of SNR and MSE is proposed.According to theoretical analysis and engineering projects,the almost-hard thresholding can be more effective to take the local deformation characteristic information than Garrote thresholding.

关 键 词：变形监测 阈值函数 信噪比 

分 类 号：P258[天文地球—测绘科学与技术]