摘要
变形表现为一种弱信号,误差呈现为强噪声。小波包分解消噪是对分解中所得到的高频部分再继续细分,因此具有更精细的消噪能力。采用最优小波包基进行信号重构的过程中,信号有一定的损失。若将消噪过程近似成拟合,拟合误差符合正态分布,可以采用假设检验法估计信号重构误差的区间,进而确定信号重构误差的不确定度,评价小波包消噪的质量。
The deformation appears as weak signal,and error appears as strong noise.The method that wavelet packet decomposes the signal to remove noise can decomposes the detail components,so it has more accurate ability of de-noising.In course of signal reconstruction with best basis of wavelet packets,the signal has some loss.Because of the error with normal distribution,the method of hypothesis test is introduced to evaluate the interval of the error caused by signal reconstruction if de-nosing approximates fitting.Then the uncertainty of the error caused by signal reconstruction is identified.The quality of de-nosing with this method is appraised.
出处
《测绘科学》
CSCD
北大核心
2007年第5期21-22,共2页
Science of Surveying and Mapping
基金
山东理工大学基金项目
中科院百人计划项目资助
关键词
小波包
变形监测
信号消噪
不确定度
wavelet packet
deformation monitoring
signal de-noising
uncertainty
