超声无损检测中的缺陷识别与噪声抑制

在传统的小波信号处理器基础上，根据解析小波变换能准确提取信号相位的特性，利用超声检测信号的相位信息，提出一种新的多缺陷识别与噪声抑制算法。该算法充分运用超声信号的时域、频率和相位信息，能检测多个具有不同频谱特性的缺陷。实验结果表明该算法不仅消噪性能好，而且提高了缺陷的纵向分辨率。

Flaw Identification and Noise Suppressing in Ultrasonic Nondestructive Testing

According to the accurate phase extraction of the analytic wavelet transform, using the phase information of an ultrasonic signal, a novel multiple flaw identification and noise suppressing algorithm is proposed in this paper on the basis of the traditional wavelet transform signal processor. The proposed technique can detect multiple flaws which exhibit different spectral characteristics by full utilizing the information of time domain, frequency and phase of the ultrasonic signal. Experimental results show that noise can be greatly reduced and the longitudinal resolution can be improved by the algorithm.