【摘要】 如何在消除噪声的同时更好保留信号中的有效信息一直是无损检测信号处理研究中的热点。尝试通过构建虚拟观测信号将单通道超声检测信号扩展为信号阵列,利用波原子方法进行降噪处理,并引入非局部均值算法进行平滑以压制波原子硬阈值滤波方式可能产生的伪吉布斯现象。用该算法对仿真信号与某碳酸钙污垢层超声检测信号进行降噪,并与小波阈值降噪进行对比。结果显示该方法对信号信噪比的提高、信号有效特征及微弱信号的保持均有较好效果。更多还原

【Abstract】 How to eliminate noise as well as preserve the effective information of signal is always a hotspot in nondestructive detection signal processing research.The single channel ultrasonic detection signal is attempted to expand to array signals by constructing a virtual observation signal,and then wave atoms method is used to realize noise reduction.The non-local means algorithm for smoothing is introduced to suppress Gibbs-like phenomena,which may be caused by the wave atoms hard threshold filtering.This algorithm is used on the simulation signal and the ultrasonic detection signal of a calcium carbonate fouling layer,and is compared with the wavelet threshold denoising.Results show that the method used in this paper has better effect on increasing the signal noise ratio,maintaining effective characteristics and weak signal.更多还原