Reconstruction of non-trivial magnetization textures from magnetic field images using neural networks

TL;DR

This paper introduces a neural network-based method to accurately reconstruct complex magnetization textures from magnetic field images, overcoming artefacts and limitations of traditional techniques.

Contribution

The authors develop a neural network approach that improves magnetization reconstruction, enabling inclusion of additional models and bounds, and reconstructs non-trivial textures previously inaccessible.

Findings

Reduces artefacts in magnetization reconstruction

Allows inclusion of additional models and bounds

Successfully reconstructs complex topological spin textures

Abstract

Spatial imaging of magnetic stray fields from magnetic materials is a useful tool for identifying the underlying magnetic configurations of the material. However, transforming the magnetic image into a magnetization image is an ill-poised problem, which can result in artefacts that limit the inferences that can be made on the material under investigation. In this work, we develop a neural network fitting approach that approximates this transformation, reducing these artefacts. Additionally, we demonstrate that this approach allows the inclusion of additional models and bounds that are not possible with traditional reconstruction methods. These advantages allow for the reconstruction of non-trivial magnetization textures with varying magnetization directions in thin-film magnets, which was not possible previously. We demonstrate this new capability by performing magnetization reconstructions on a variety of topological spin textures.