Ultrasonic signals parameters estimation based on differential evolution

Many techniques, such as parameter extraction techniques are used in ultrasonic signal processing. In this paper, the matching pursuit decomposition (MP) and differential evolution (DE) have been tested for parameters estimation, denoising and compression of ultrasonic echoes. The method is based on the MP algorithm who does the decomposition of the ultrasonic signal into elementary functions (atoms) which they make a good representation of this signal. Those elementary functions are selected from a Gabor dictionary and their parameters are estimated by differential evolution.