基于空域相关滤波的一种新的小波降噪算法  

A New Wavelet De-noising Method Based on Spatial Correlation Filtering

作　　者：马强[1] 王洋[1] 刘海萍[1] 

机构地区：[1]91245部队

出　　处：《计算机与数字工程》2014年第11期2045-2048,共4页Computer & Digital Engineering

摘　　要：由于信号与噪声在小波域具有不同的表现,空域相关滤波算法通过取相邻尺度的小波系数相乘进行相关运算,使得在保留信号细节的同时而抑制噪声。但空域相关滤波算法没有明确给出在每层尺度上提取信号细节的次数。论文主要针对此缺点,提出一种新的算法。通过构造损失函数,使得当损失函数最小时,确定出最佳的降噪信号。仿真计算表明该算法取得了较好效果。Since the signal and noise have different performances in wavelet domain, spatial correlation filtering algorithm con- ducts related operations by multiplying adjacent scale wavelet coefficients, which can be able to keep details of signal and suppress noise at the same time. But this algorithm doesn＇t give clear times of extracting the signal details on each scale. This paper mainly aims at this drawback and proposes a new method. Through constructing loss function, the best de-noising signal is determined when this function is smallest. The simulation calculation shows that this method achieves Rood effect.

关 键 词：小波 降噪 滤波 空域相关 

分 类 号：TP301.6[自动化与计算机技术—计算机系统结构]