盲小波算法在遥感图像去噪中的应用  

The Application of Blind-Wavelet Algorithm on De-noising of the Remote Sensing Data

作　　者：吴海洋[1,2] 周仲礼[1,2] 曾凡霞[1,2] 胥德平[1,2] 

机构地区：[1]成都理工大学,成都610059 [2]数学地质四川省重点实验室,成都610059

出　　处：《软件》2012年第1期9-11,共3页Software

基　　金：国家自然科学青年基金项目(409904034)资助;数学地质四川省高校重点实验室开放基金项目(SCSXDZ2009016)资助

摘　　要：根据盲信号分离原理和小波分析,提出了一种遥感图像去噪的盲小波算法,首先将遥感图像的个信号进行同深度小波分解,得到不同信号相应深度的小波系数和尺度系数,然后将小波系数进行软阈值法处理,并进一步对不同信号的同深度的小波系数和尺度系数进行盲分离,并提取与源信号相关的信号,最后通过信号重构估计源信号。这种将小波分析和盲信号分离技术有机结合的方法能够有效的消除遥感图像的噪声。通过对实际遥感图像的处理,并与其他去噪技术相比较,利用盲小波算法得到的结果更为理想。Based on blind source separation and wavelet analysis, a Blind-Wavelet algorithm that to eliminate remote sensing data noise is proposed. This new method contains the follow three main steps： firstly, the multi-channel seismic signals are decomposed into multi-level wavelet to obtain the scaling coefficients and the multi-level wavelet coefficients that are processed by using soft threshold method at the same time; secondly, it is need to use blind source separation to process all the scaling coefficients and all the same depth wavelet coefficient in the correct order, meanwhile the different frequency band signals that are closely related to the source signals are extracted through an appropriate manner; finally, the source signal is obtained by the reconstruction signal of wavelet transform. The new method is the organic combination of blind source separation and wavelet analysis can effectively eliminate the noise of remote sensing data. Compare to the other de-noising methods, the new de-noising method obtained more satisfactory results through the real remote sens- ing data processing.

关 键 词：遥感图像去噪 小波分析 盲信号分离 最小噪声分离 

分 类 号：TP751[自动化与计算机技术—检测技术与自动化装置]