The dispersive interferometry provides an instantaneous surface measurement in a single camera frame, making it resistant to environmental disturbances and ideal for in-process surface metrology. It also benefits from the extended measurement ranges in both depth and lateral directions by incorporating hyperspectral imaging technology and cylindrical beam illumination, respectively. This paper reports on an in-house developed cylindrical lens-based dispersive interferometer for high-accuracy surface inspection, particularly for structured surfaces. The obtained spectral interferogram is analyzed using the fringe order algorithm, in which the phase slope method is used to calculate the initial height to resolve the fringe order ambiguity and eventually an improved height value can be obtained using the exacted phase of a single wavelength. Experiments demonstrate that the measurement noise of the developed interferometry system is less than 1 nm within the measurement range. A brass step sample made by a diamond turning machine was measured and the experimental results closely align with those given by the commercial white light interferometer -Talysurf CCI 3000.
Surface reconstruction method plays an important role in many engineering fields. It is an imperative procedure to carry out surface reconstruction from measurement data in reverse engineering, which is complicated with the presence of outliers. To achieve better accuracy and robustness of reconstruction, an improved moving total least squares (MTLS) algorithm based on k-means clustering called KMTLS method is proposed in this article. KMTLS adjusts the weights of discrete points within the support domain by adopting a two-step fitting procedure. Firstly, ordinary least squares (OLS) method is adopted to obtain the pre-fitting result and calculate the residuals as the input of k-means clustering. In kmeans clustering, abnormal nodes are classified into one cluster and a weight function based on clustering information is introduced to deal with these nodes. Secondly, based on the compact weight function in MTLS and the weight obtained in the pre-fitting procedure, weighted total least squares method is conducted to determine the final estimated value. The process of detecting outliers is automatic without setting threshold artificially. The experiment shows that KMTLS has great robustness to outliers.