Nondestructive evaluation (NDE) is the process of detecting, locating, characterizing and sizing of an anomaly in an engineering material. There are many methods of performing NDE of which ultrasound is a widely used technique. Ultrasound has been used for several decades as a tool for nondestructive evaluation and material characterization and has emerged as a powerful method of analysis with the on going computer revolution. Many powerful and computationally intensive methods and algorithms have become feasible with the availability of very fast and not-too-expensive computers. Analog-to-digital conversion and digital signal processing have become common in ultrasonic signal analysis for both nondestructive evaluation and noninvasive diagnosis. Some of the signal analysis techniques in use today are (1) deconvolutions [1], (2) Homomorphic signal processing techniques such as cepstrum analysis [1], (3) feature extraction and feature classification techniques [2] and (4) artificial neural networks [3, 4]. These signal analysis techniques have been shown to be very effective for all the phases of NDE ranging from detection to characterization and sizing.