Quantitative spherometry through all-optical differentiation with a computing metasurface
Phys. Rev. A 113, 023507 – Published 5 February, 2026
DOI: https://doi.org/10.1103/qgp6-ztgq
Abstract
The precise, nondestructive quantification of local surface curvature is paramount for advancements in precision manufacturing, biomedical engineering, and industrial quality control. However, conventional techniques, including contact profilometry and optical interferometry, are constrained by potential surface damage, limited adaptability to complex geometries, and resolution trade-offs. To address these limitations, we introduce and experimentally demonstrate a noncontact spherometer based on an optical computing metasurface, enabling high-precision local curvature metrology at the micrometer level. This modality integrates all-optical object recognition for edge detection of intensity objects with quantitative phase imaging for the three-dimensional reconstruction of phase objects. Experimental validation on calibrated targets, including a standard steel sphere and a convex lens, demonstrates that the experiment exhibits high-precision characteristics. This work establishes metasurface-based optical computing as a versatile and robust platform for high-precision metrology, opening new avenues for in situ surface characterization across multiple disciplines.
