Application of a two-dimensional phase unwrapping algorithm to lubricant film thickness measurement

In the precise characterization of microscopic surface morphology, phase unwrapping constitutes a critical step for achieving quantitative reconstruction in interferometric measurements. In lubricating oil-film contact imaging, phase unwrapping algorithms are frequently affected by noise interference, edge blurring, and dynamic operating conditions, which can lead to phase discontinuities and a limited effective unwrapping range. These issues consequently degrade the continuity and accuracy of the reconstructed results. To improve the robustness of phase unwrapping and the accuracy of oil-film thickness reconstruction, this study develops an improved unwrapping method based on the Sorting by Reliability following a Non-Continuous Path (SRNCP) framework for lubricating oil-film interferometric images. By combining phase gradients with local noise estimation to construct a composite quality map and integrating it into the non-continuous path-guided unwrapping process, the proposed method enhances phase unwrapping stability and reconstruction accuracy under complex interferometric conditions. Comparative experiments demonstrate that the improved algorithm achieves significantly enhanced performance in oil film thickness reconstruction compared to traditional methods. Finally, applying the improved unwrapping algorithm to practical lubricating oil film thickness detection further validates its engineering applicability under complex operating conditions.