Statistical data analysis and signal processing techniques for imaging and non-destructive testing

The course introduces some statistical data analysis and signal processing techniques for imaging and non-destructive testing applications which all rely on theleast-squares theory.By starting with solving the linear regressionproblem with noisy data, the least-squares theory will be introduced and its relationship with the pseudo-inverse matrixillustrated.The pseudo-inverse approach will be then used to solve generic polynomial fitting, regression and non-linearfitting through iterative algorithms.The basic principles of inverse theory and deconvolution will be suddenly introduced and then the course will focus on signal processing applications, all developed for dealing with optimal deconvolution
and linear system characterization in noisy environment.Wiener filter theory, matched filter theory, pulse-compressionand ARMA modelling will be presented and numerical exercises will be done by the students during the course.

Number of CFU: 3

TEAMS access code: vjk7o1y