Characteristic extraction of ultrasonic detection of bonding interface of thin composite plate based on wavelet transformation

This paper studies an algorithm of characteristic extraction of ultrasonic detection based on wavelet transformation. The object of ultrasonic detection is thin composite plate. At first, we discuss characteristic and composing principle of the detail of echo signal, based on this a math model is designed. And then, wavelet transformation is used to process echo signal model and defect signal model of different viscid quality, we extract characteristic about energy and dimension in time-frequency domain. Simulation by the Matlab proves that the algorithm can recognize different echo signals effectively, and it is helpful to quantitative detection and recognition.