Bayesian analysis of random tessellations

We consider a 3D point pattern (pores in aluminium) which apart from Gaussian noise is described by a Poisson process defined on the edges of a spatial Poisson-Voronoi tessellation. Simulation-based Bayesian methods are applied with the aim of estimating various characteristics of the unobserved Voronoi cells. Cooperation with Øivind Skare, Norwegian Computing Center, Norway, and Eva Vedel, Aarhus University, Denmark Supported by MaPhySto and the Danish Natural Science Council.