Spectral and 3D motifs identification of anisotropic topographical components. Analysis and filtering of anisotropic patterns by morphological rose approach

Abstract

In this paper we show two new developments in the three-dimensional characterisation of rough surfaces. The basic idea is to consider roughness as a combination of components defined by the roughness amplitude, wave length, local and overall direction of various components. The anisotropic topography is dealt with in two ways: • - the anisotropy between form, waviness and roughness; • - the local anisotropy of three-dimensional motifs. The directional parameter is used both to identify and separate the anisotropic components by appropriate anisotopic filtering and by a complete characterisation of surface motifs. This allows the decomposition of the topography into local motifs with a representation similar to the two-dimensional Fourier transform, where the direction represents the topographic phase in the form of a morphological rose.