Abstract
High-resolution optical microscopy, particularly super-resolution localization microscopy, requires precise real-time drift correction to maintain constant focus at nanoscale precision during the prolonged data acquisition. Existing methods, such as fiducial marker tracking, reflection monitoring, and bright-field image correlation, each provide certain advantages but are limited in their broad applicability. In this work, a versatile and robust drift correction technique is presented for single-molecule localization-based super-resolution microscopy. It is based on the displacement analysis of bright-field image features of the specimen with oblique illumination. By leveraging the monotonic relationship between the displacement of image features and axial positions, this method can precisely measure the drift of the imaging system in real-time with sub-nanometer precision in all three dimensions, over a broad axial range, and for various samples, including those with closely matched refractive indices. The performance of this method is validated against conventional marker-assisted techniques and demonstrates its high precision in super-resolution imaging across various biological samples. This method paves the way for fully automated drift-free super-resolution imaging systems.