High speed design and performance evaluation of frequency-diverse orthogonal transforms for ultrasonic imaging applications

Ultrasonic flaw detection in the presence of high scattering grain noise is a significant problem in the non-destructive evaluation (NDE) of materials. In this work, signal processing methods for ultrasonic flaw detection that incorporate discrete cosine transform (DCT), and discrete Walsh-Hadamard transform (WHT) will be presented. These are orthogonal transforms that are suitable candidates for integration in a system-on-chip (SoC) architecture design where computational complexity may be critical. The flaw detection systems for DCT and WHT are implemented in a way that is similar to the conventional split spectrum processing (SSP) methods. Experimental data results demonstrate that these methods are as effective as the SSP method for flaw detection in presence of high scattering grain noise. Efficient designs to optimize the hardware architectures for DCT and WHT based flaw detection applications are also discussed.