Subaperture stitching algorithms: A comparison

With the research focus of subaperture stitching interferometry shifting from flat wavefronts to aspheric ones, a variety of algorithms for stitching optimization have been proposed. We try to category and compare the algorithms in this paper by their modeling of misalignment-induced subaperture aberrations which are of low orders. A simple way is to relate the aberrations to misalignment by linear approximation with small angle assumption. But it can not exactly model the induced aberrations of aspheres. In general, the induced aberrations can be fitted to free polynomials and then removed from subaperture measurements. However, it is at the risk of mixing up the surface error and the induced aberrations. The misalignment actually introduces different terms of aberrations with certain proportions. The interrelation is then determined through analytical modeling or ray tracing. The analytical model-based algorithm and the ray tracing-based algorithm both are tedious, aperture shape-related and surface type-specific. While the configuration space-based algorithm we proposed numerically calculates the surface height change under rigid body transformation, it is generally applicable to various surface types and different aperture shapes. Simulations and experiments are presented to compare the stitching results when different algorithms are applied to null cylindrical subapertures measured with a computer generated hologram. The configuration space-based algorithm shows superior flexibility and accuracy.