Issue No. 06 - June (2013 vol. 35)
Patrick Héas , INRIA, Rennes
Cédric Herzet , INRIA, Rennes
Etienne Mémin , INRIA, IRSTEA, Rennes
Dominique Heitz , IRSTEA, Rennes
Pablo D. Mininni , University of Buenos-Aires, Buenos Aires and National Center for Atmospheric Research, Buenos Aires
Based on physical laws describing the multiscale structure of turbulent flows, this paper proposes a regularizer for fluid motion estimation from an image sequence. Regularization is achieved by imposing some scale invariance property between histograms of motion increments computed at different scales. By reformulating this problem from a Bayesian perspective, an algorithm is proposed to jointly estimate motion, regularization hyperparameters, and to select the most likely physical prior among a set of models. Hyperparameter and model inference are conducted by posterior maximization, obtained by marginalizing out non--Gaussian motion variables. The Bayesian estimator is assessed on several image sequences depicting synthetic and real turbulent fluid flows. Results obtained with the proposed approach exceed the state-of-the-art results in fluid flow estimation.
Bayesian methods, Optimization, Vectors, Estimation, Motion estimation, Computational modeling, Optical imaging, Bayesian model selection, Optic flow, turbulence, robust estimation, constrained optimization
P. D. Mininni, E. Mémin, D. Heitz, P. Héas and C. Herzet, "Bayesian Estimation of Turbulent Motion," in IEEE Transactions on Pattern Analysis & Machine Intelligence, vol. 35, no. , pp. 1343-1356, 2013.