A simulation method to estimate task-specific uncertainty in 3D microscopy

Abstract

Traceability in micro-metrology requires an infrastructure of accredited metrology institutes, effective performance verification procedures, and task specific uncertainty estimation. Focusing on the latter, this paper proposes an approach for the task specific uncertainty estimation based on simulation for a generic 3D microscope. The proposed simulation approach is based on the identification and a successive parameter estimation of an empirical model of measured points. The model simulates the probing error of the 3D microscope based on a Gaussian process model, thus including the correlation among close points. Parameters for the error simulation are estimated by a deep analysis of error sources of the 3D microscope. Validations of the proposed simulation approach are carried out in the case of focus variation microscopy (FVM), considering several case studies. The procedure proposed in the ISO/TS 15530-4 standard are applied for validation.