vix.ing · top · new · best · stats · spec

Fast Scanning Probe Microscopy via Machine Learning: Non-rectangular\n scans with compressed sensing and Gaussian process optimization

2020/04/23 by Kyle P. Kelley, Kelley, Kyle P., Maxim Ziatdinov +13 · 2 citations
Physics and Astronomy · Engineering · #Force Microscopy Techniques and Applications #Near-Field Optical Microscopy #Piezoelectric Actuators and Control

paper · pdf · doi:10.48550/arxiv.2004.11817

Abstract

Fast scanning probe microscopy enabled via machine learning allows for a\nbroad range of nanoscale, temporally resolved physics to be uncovered. However,\nsuch examples for functional imaging are few in number. Here, using\npiezoresponse force microscopy (PFM) as a model application, we demonstrate a\nfactor of 5.8 improvement in imaging rate using a combination of sparse spiral\nscanning with compressive sensing and Gaussian processing reconstruction. It is\nfound that even extremely sparse scans offer strong reconstructions with less\nthan 6 % error for Gaussian processing reconstructions. Further, we analyze the\nerror associated with each reconstructive technique per reconstruction\niteration finding the error is similar past approximately 15 iterations, while\nat initial iterations Gaussian processing outperforms compressive sensing. This\nstudy highlights the capabilities of reconstruction techniques when applied to\nsparse data, particularly sparse spiral PFM scans, with broad applications in\nscanning probe and electron microscopies.\n

Cited by

Related