基于小波脊的超声信号处理  

Ultrasonic Signal Processing Based on Wavelet Ridge

作　　者：张广明[1] 管鸣宇[1] 王裕文[1] 谭玉山[1] 

机构地区：[1]西安交通大学

出　　处：《西安交通大学学报》1999年第7期94-98,共5页Journal of Xi'an Jiaotong University

摘　　要：根据超声信号通常是单频载波脉冲信号，具有渐近信号的特性，提出了一种基于小波脊的超声信号处理算法．首先利用渐近小波变换提取超声信号的小波脊，计算小波脊的移动熵，然后通过移动熵实现缺陷信号与噪声的分离和缺陷定位，最后利用小波反变换重构出缺陷信号．该算法充分利用了超声信号的时域、频域和相位信息，不仅消噪性能好。Ultrasonics are used as the single carrier frequency impulse for characterising asymptotic signals. An algorithm is developed for processing the ultrasonic signal by first extracting the wavelet ridge using the asymptotic wavelet transform and calculating the moving entropy which detects the flaw location and separates the flaw signal and noise. Inverse wavelet transform is then applied to reconstruct the flaw singal. The time domain, frequency domain and phase of ultrasonic signal are fully utilized such that noise is discarded and flaw is located with accuracy.

关 键 词：小波脊 超声检测 信号处理 无损检测 

分 类 号：TH878.2[机械工程—仪器科学与技术] TG115.285[机械工程—精密仪器及机械]