Assessing planar asymmetries in shipbuilding from point clouds

This paper introduces a statistical methodology to identify planar asymmetries in surfaces. An algorithm based on bivariate kernel smoothers is used to estimate non-parametrically the surface from a point cloud. Also bootstrap-based procedures are proposed for testing a null hypothesis establishing that the surface has planar symmetry. After checking the validity of the proposed methodology with simulated data, it was applied to study symmetries in a boat hull.