Universal Drift Correction for Multidimensional Scanning Microscopy

Sangjoon Lee; William Millsaps; Dasol Yoon; Caitlyn Obrero; Guoliang Hu; Corrie Barnes; Cedric Lim; Andrew Barnum

cond-mat.mtrl-sciSep 25, 2026

Universal Drift Correction for Multidimensional Scanning Microscopy

Authors: Sangjoon Lee, William Millsaps, Dasol Yoon, Caitlyn Obrero, Guoliang Hu, Corrie Barnes, Cedric Lim, Andrew Barnum,

Organizations: Department of Materials Science and Engineering, Stanford University, Stanford, CA 94305, USA · Stanford Nano Shared Facilities, Stanford University, Stanford, CA 94305, USA

Abstract

In scanning microscopy, drift causes the specimen to be sampled at positions displaced from the nominal probe positions. This displacement alters the spatial assignment of the recorded signals and biases quantitative measurements across two-dimensional imaging, channel-resolved spectroscopic mapping, and scan-position-resolved diffraction analysis. Here, we extend orthogonal-scan drift correction from 2D images to spectrum images and diffraction datasets. We demonstrate how to recover probe positions using either differently oriented multidimensional scans or structural reference images. The recovered positions are used either to resample the multidimensional data onto a regular grid or to assign each recorded signal to its corrected coordinate. Our method combines affine and non-rigid correction, requires no prior structural model, and is implemented as open-source, GPU-accelerated software that reduces processing times by two to three orders of magnitude, enabling routine and automated drift correction for quantitative multidimensional microscopy.

Figures & tables

Figure 1: Universal drift correction by recovering corrected scan-line origins for each rastered acquisition. (a) Differing-angle scans record distinct distortions of the same specimen where affine and non-rigid correction stages can recover the probe positions sampled on the specimen. For each scan, the recovered origins are reused for 2D images, channel-resolved multidimensional data, and scan-position-resolved diffraction data. (b) Knot representation of scan geometry, from the initial scan-line grid to affine correction with one knot per scan line and non-rigid correction with three knots per scan line.

Figure 2: Affine drift correction of a high-angle annular dark-field (HAADF) image of silicon, using a $0^{\circ}$ / $90^{\circ}$ scan pair collected from the same specimen region: (a) $0^{\circ}$ scan; (b) $90^{\circ}$ scan; (c) the combined scan before correction; (d) the $0^{\circ}$ scan after correction; (e) the $90^{\circ}$ scan after correction, with red lines in (d, e) marking the recovered scan-line origins; and (f) the combined scan after only affine correction.

Figure 3: Affine and nonrigid drift correction of WS 2 images. Columns, from left to right: raw $0^{\circ}$ image, raw $90^{\circ}$ image, affine-corrected image, nonrigid-corrected image assuming no drift within each fast-scan line, and nonrigid-corrected images using two and three control points per fast-scan line. Columns 3–6 combine the two raw images, improving the signal-to-noise ratio. Rows, from top to bottom: processed image, identified A sites, nearest-neighbor bond-vector distributions, and collapsed bond-vector distributions after centroid subtraction.

Figure 4: Center-out non-rigid correction in $\bf{WSe_{2}}$ . The raw frame exhibits extreme drift that is not fully corrected by standard affine and non-rigid refinement. The center-out procedure prevents unit-cell skips.

Figure 5: Drift correction across sample types. Rows: silicon, SrTiO 3 , a WS 2 monolayer, and Co 3 O 4 . Columns: the $0^{\circ}$ and $90^{\circ}$ scans; magenta/green RGB combined views of the two scans before correction, after affine correction, and after non-rigid correction, where misregistration appears as color separation and alignment as white; and the final drift-corrected combined image. All panels of a row share one orientation and field of view, and every panel is rendered by the same pipeline, so the columns differ only by the correction stage.

Figure 6: Drift correction of a SrTiO 3 energy-dispersive X-ray spectroscopy (XEDS) spectrum image. (a, b) The $0^{\circ}$ XEDS HAADF before and after correction; (c) the corrected $0^{\circ}$ / $90^{\circ}$ HAADF reference. (d, e) The Ti K map (integrated over $4.30$ – $4.80$ keV) before and after correction; (f) the Ti K (magenta) and Sr L ( $1.70$ – $1.95$ keV, green) composite on the corrected HAADF.

Figure 7: Four-dimensional (4D) scan-drift correction of gold nanoparticles. (a, b) Virtual bright-field images from the $0^{\circ}$ and $90^{\circ}$ acquisitions, with (b) rotated into the $0^{\circ}$ specimen frame. (c) Uncorrected combined image. (d, e) Corrected virtual bright-field images from the two acquisitions. (f) Corrected combined image. Insets in (c) and (f) show the individual $0^{\circ}$ and $90^{\circ}$ diffraction patterns nearest the selected positions, without spatial averaging or detector-pixel interpolation. The virtual bright-field images and diffraction patterns use shared display ranges.

Microscopy Image Analysis

Microscopy image analysis uses computational methods to detect, segment, track, reconstruct, and quantify structures in scientific images.

See all 198 papers

Multitask Scanning Probe Microscopy
Aug 10, 2026Aditya Raghavan, Yu Liu, Ian Mercer +2Active Learning Multi-Task Learning
University of Tennessee, Knoxville, TN, USA · 2Pennsylvania State University, University Park, PA, USA
Novelty-Driven Target-Space Discovery in Automated Electron and Scanning Probe Microscopy
Mar 17, 2026Utkarsh Pratiush, Kamyar Barakati, Boris N. Slautin +4Deep Kernel Learning Scanning Transmission Electron Microscopy
Department of Materials Science and Engineering, University of Tennessee, Knoxville, TN 37996, USA · Department of Chemistry, University of Washington, Seattle, WA 98195, USA · Pacific Northwest National Laboratory, Richland, WA 99354, USA
$\texttt{DisMorph}$ : learning to disentangle technical distortions from true biological change
Aug 8, 2026Jingru Fu, Kathleen E. Larson, Douglas N. Greve +2Longitudinal Medical Image Analysis Medical Image Registration
Athinoula A. Martinos Center for Biomedical Imaging, Charlestown, USA · Department of Radiology, Massachusetts General Hospital, Boston, USA · Department of Radiology, Harvard Medical School, Boston, USA

Aug 10, 2026J/K move · Enter open · S save

Multitask Scanning Probe Microscopy

Aditya Raghavan, Yu Liu, Ian Mercer, JP Maria+1

University of Tennessee, Knoxville, TN, USA · 2Pennsylvania State University, University Park, PA, USA

Active Learning Multi-Task Learning Autonomous Experimentation

Open paper AskPDF arXiv

Scanning probe microscopy provides nanoscale access to structural, electrical, electromechanical, magnetic, and mechanical properties of materials. Its increasing use for wafer-scale characterization and combinatorial materials exploration creates a need to distribute measurements efficiently across large spatial domains. This is particularly important when available modalities differ in acquisition time and potential for tip and sample damage, making exhaustive multimodal mapping over spatial grids impractical. Here, we demonstrate multitask scanning probe microscopy, a live, closed-loop workflow in which a multitask Gaussian process learns spatial and cross-modal relationships and autonomously selects both the next measurement location and the next experimental protocol. The approach is implemented on an automated large-sample atomic force microscope and demonstrated on a composition-spread AlScN wafer using tapping-mode and Dual AC Resonance Tracking (DART) measurements. Paired initial measurements establish the relation between the tasks, after which noncoincident measurements are used to update both response landscapes. The resulting workflow extends active learning in scanning probe microscopy from spatial sampling to autonomous allocation of measurement modalities and provides a basis for combining rapid, weakly perturbative imaging with slower contact, electrical, electromechanical, magnetic, or spectroscopic measurements.

Abstract

Microscopy Image Analysis

Explore similar work

Multitask Scanning Probe Microscopy

Novelty-Driven Target-Space Discovery in Automated Electron and Scanning Probe Microscopy

DisMorph\texttt{DisMorph}DisMorph: learning to disentangle technical distortions from true biological change

$\texttt{DisMorph}$ : learning to disentangle technical distortions from true biological change