Ultrasonic flaw detection using discrete wavelet transform for NDE applications

In this work, we analyze signal decomposition properties of discrete wavelet transform (DWT) for enhanced ultrasonic flaw detection. In wavelet signal decomposition, a collection of time-frequency representations of the signal with different resolutions is obtained. DWT allows to utilize both time and frequency domain information for compacting and decorrelating the flaw echo from clutter echoes. In this paper, we present the performance analysis of different wavelet kernels with respect to ultrasonic NDE applications and develop the wavelet selection criteria for optimal flaw detection. Experimental results indicate that DWT based flaw detection algorithms offer flaw-to-clutter ratio enhancement of 5-12 dB when the measured flaw-to-clutter ratio is 0 dB or less. DWT flaw detection system can be implemented efficiently for real time applications using reconfigurable architecture and lifting scheme.