Real-time Rigid-body Visual Tracking in a Scanning Electron Microscope

Abstract

Robotics continues to provide researchers with an increasing ability to interact with objects at the nanoscale. As microrobotic and nanorobotic technologies mature, more interest is given to computer-assisted or automated approaches to manipulation. Although actuators are currently available that enable displacement resolutions in the subnanometer range, improvements in feedback technologies have not kept pace. Thus, many actuators that are capable of performing nanometer displacements are limited in automated tasks by the lack of suitable feedback mechanisms. This paper proposes the use of a rigid-model-based method for end-effector tracking in a scanning electron microscope to aid in enabling more precise automated manipulations and measurements. These models allow the system to leverage domain-specific knowledge to improve performance in a challenging tracking environment.