Sparse representation of a force spectrum and a force-volume image
Résumé
We present a new method for analyzing AFM data, relying on the concept of sparse representation. A signal is called sparse if it can be described by a very limited set of parameters. For AFM spectra, the sparse description is based on the discontinuities (jumps and derivatives) embedded in the data. The communication aims at presenting a very recent signal processing method to analyze sparse signals (denoising and extraction of patterns). The method is introduced and some results on real data are displayed, showing that patterns can be extracted in a very accurate fashion.