基于自适应阈值的小波消噪算法研究及应用  

Study of Wavelet De-noising Algorithm and Its Application Based on Adaptive Thresholding Method

作　　者：毛海杰[1] 陈晓辉[1] 李炜[1] 

机构地区：[1]兰州理工大学电气工程与信息工程学院,甘肃兰州730050

出　　处：《甘肃科学学报》2008年第2期116-119,共4页Journal of Gansu Sciences

基　　金：教育部"春晖计划"项目(Z2005-1-62001);兰州理工大学重点特色梯队建设项目(0915)

摘　　要：基于小波多尺度分解的阈值方法是一种有效的信号去噪方法.应用最为广泛的是Donoho阈值法,但由于其阈值为全局阈值,不能在每级尺度上将信号与噪声做最大分离,去噪效果并不理想.在分析Donoho硬阈值和软阈值方法特点的基础上,提出了一种新的自适应阈值消噪方法,结合了硬阈值和软阈值方法的优点,仿真试验和工程实例均证明了该方法优于传统的阈值消噪方法.An effective method to reduce the noise in signals is based on wavelet multiscale decomposition. The thresholding de-noising proposed by Donoho is a standard wavelet de-noising method. But it has some disadvantages. Based on Donoho＇s method, an improved wavelet de-noising method with adaptive thresholding is given, which has both advantages of hard-thresholding and soft-thresholding. The simulation experiment and project practice indicate that the proposed method is better than traditional wavelet thresholding de-noising methods.

关 键 词：小波多尺度分解 阈值消噪 管道泄漏 

分 类 号：TN911.23[电子电信—通信与信息系统]