Approximate extraction of late-time returns via morphological component analysis

A fundamental challenge in acoustic data processing is to separate a measured time series into relevant phenomenological components. given measurement typically assumed be an additive mixture of myriad signals plus noise whose separation forms ill-posed inverse problem. In the setting sensing elastic objects using active sonar, we wish early-time returns (e.g., from object's exterior geometry) late-time caused by or compressional wave coupling. Under framework Morphological Component Analysis (MCA), compare two models short-duration and long-duration responses as proxy for returns. Results are computed Stanton's cylinder model well on experimental taken in-Air circular Synthetic Aperture Sonar (AirSAS) system, separated formed imagery. We find that MCA can used early both cases without use time-gating. The process demonstrated robust compatible with AirSAS image reconstruction. best results obtained flexible, but computationally intensive, frame based signal model, while faster Fourier Transform method shown have competitive performance.