Ultrasonic flaw detection using sparse representation for failure analysis of next generation microelectronic packages

In this paper, a sparse Bayesian learning-based ultrasonic signal processing method is designed to detect and characterize flaws in thin multilayer structures, where the reflectivity is sparse. Sparse Bayesian learning is first used to decompose an ultrasonic signal into sparse signal representations over an overcomplete dictionary that is learnt from a training data set in advance. Ultrasonic flaw detection is then carried out on the basis of the sparse signal representations. The performance of the proposed method is experimentally verified using ultrasonic traces acquired from microelectronic packages by a scanning acoustic microscope.