A flaw echo detection method based on S-transformation and time-frequency spectrum spatial filtering

Xiang ZENG,Xuelei WU,Xiaojun ZHOU,Chenlong YANG,Yuechao CHEN

DOI: https://doi.org/10.13465/j.cnki.jvs.2017.22.006

2017-01-01

Abstract:Detection of flaw echoes is an important task in ultrasonic testing.In order to eliminate the effects of noise and detect echoes correctly,a signal processing method based on the S-transformation (ST) and the time-frequency spectrum (TFS) spatial filtering was proposed.Based on the Gaussian echo model,the relationships between the time of arrival (TOA) together with the center frequency (CF) and ST TFS were discussed.The reasonableness of the method using the extremum of ST TFS amplitude matrix for TOA and CF extraction were illustrated.To detect the echoes,firstly the ST was performed on the original signal.Then the dual-threshold were determined by the Maximum Entropy Thresholding method adaptively,and the spatial filtering was performed on ST TFS for denoising.The TOA and CF could be extracted from the denoised region.Finally,the inverse ST was performed on the denoised TFS and the signal with clear flaw echoes were gained.Simulation results show that using ST and TFS spatial filtering can remove noise and detect echoes effectively.And the experimental results of bar specimen also show the effectiveness of the method in flaw detection.