“…In order to apply magnetic microscopy to paleomagnetic studies, it is necessary to recover from the magnetic images a large number of individual magnetic moments, corresponding to at least tens of thousands of stable fine-grained grains (< 1 µm), in order to provide statistical significance to the remanence vector (e.g., Berndt et al, 2016). Nowadays, with the development of magnetic microscopy techniques, this task is no longer limited by the resolution of magnetic microscopes (de Groot et al, 2018;Fu et al, 2020;Glenn et al, 2017;Lima et al, 2014;Weiss et al, 2007), but essentially by the intrinsic problem presented by the ambiguity in the inversion of potential field data (Barbosa and Silva, 2011;de Groot et al, 2021;Oliveira Jr. et al, 2015), and ultimately by the lack of a fast and automated way to recover such a large number of individual magnetic moments from a set of magnetic images (Cortés-Ortuño et al, 2022;Lima and Weiss, 2009;Lima et al, 2013).…”