Influence of Calibration Artefact Geometry on Pixel-Based Dimensional Measurements in Optical Microscopy

Calibration is a critical step in image-based dimensional metrology because the calibration coefficient directly determines the conversion between image pixels and physical dimensions. While numerous artefact models are available for optical microscopes, the influence of calibration artefact geometry on measurement performance has not been sufficiently investigated. This study evaluates the effect of linear and circular calibration artefacts on the calibration and verification of a pixel-based optical measuring microscope. Calibration coefficients were determined using linear and circular artefacts with nominal dimensions of 1.000 mm, 0.500/0.508 mm, and 0.200 mm at four objective magnifications (1×–4×). The calibrated microscope was subsequently verified using independent Grade 1 ceramic gauge blocks with nominal dimensions of 1.00504 mm, 1.29999 mm, and 1.50007 mm. Statistical evaluation was performed using Kruskal–Wallis testing, Dunn–Šidák post hoc analysis, effect size assessment, and measurement uncertainty analysis. The results demonstrated that calibration artefact geometry influences the obtained calibration coefficients and can produce statistically significant differences in dimensional measurements. However, uncertainty analysis revealed that the dominant contribution to the uncertainty budget originated from the microscope maximum permissible error, whereas the contribution of calibration artefact geometry remained comparatively small. All verification measurements remained within the declared metrological performance of the microscope. For the investigated microscope and experimental conditions, both calibration target geometries provided satisfactory calibration performance for the investigated microscope and measurement conditions.