Micrometric Probe and Sample Position Measurement in Scanning Probe Microscopy Through Computer Vision

Abstract

This work presents a computer vision-based methodology for precise, dynamic probe-sample distance measurement in scanning probe microscopy. The technique is tested by scanning a representative planar microwave probe and monitoring its position in three dimensions using a stereoscopic optical microscope. The results demonstrate that, through triangulation, the spatial resolution of the three-dimensional system surpasses that of the individual optical microscopes. This paper also introduces an approach that addresses the challenges of camera calibration and the limited number of feature matches between images, using augmented reality tags (ARTags) and Kalman filters to ensure process continuity and robustness. A notable feature of the methodology is its ability to estimate distances without relying on specific sample characteristics or direct contact with the sample surface. The proposed algorithms
are scalable, allowing for the generation of partial or complete reconstructions of micrometric.