Joint Estimation of Volume and Velocity in TomoPIV
Résumé
A novel formulation of the nexus between instantaneous volumetric reconstruction and velocity retrieval with respect to Tomographic PIV context is presented. In the first part of this paper we relate to the state-of-the-art paradigm and introduce a sequential estimation approach which computes velocity fields out of consecutive 3D intensity volumetric distributions. The latter are solutions of '1-minimization problems which are solved by basis pursuit algorithms. We then express the velocity estimation in a computer vision framework, as a scene flow computation problem. In the second part of the paper we formulate an innovative penalty criterion as an intensity conservation function with respect to both the fluid trajectory and the static volumetric intensity distributions, which allows us to coherently optimize the 3D intensity consecutive topologies and their related flow field with regard to their shared structure. The performance is assessed comparing the results from both sequential and joint models with respect to a synthetic tomographic imaging system.
Origine : Fichiers produits par l'(les) auteur(s)