Event-based vision in magneto-optic Kerr effect microscopy

Magneto-optic Kerr effect (MOKE) microscopy is a widely used technique for observation and characterization of microscopic magnetic structures. While being efficient and easy-to-use, current commercial MOKE microscopes are not superb in time resolution, limited by the frame rate of the camera. Here, we introduce a revolutionary sensor, namely, the event camera, as a convenient add-on to traditional MOKE microscopy and explore the potential applications of event-based vision in research areas using MOKE microscopy. We use the frame stacking method to improve visibility to human eyes in generated slow motion videos. We perform a proof-of-principle feedback control experiment using the event-based vision data and characterize the overall latency of the feedback loop as short as 25 ms with our current prototype. Finally, we discuss the limitations of current event cameras in MOKE microscopy as well.

Automated summary

The paper introduces a revolutionary event camera sensor as an add-on to traditional MOKE microscopy and explores the potential applications of event-based vision in this field. The use of frame stacking method improves the visibility of generated slow motion videos to human eyes, and a proof-of-principle feedback control experiment is performed using event-based vision data.