Abstract: A very powerful technique for solving the kind of inverse problems that often arise in the processing of fringe pattern images is based on Bayesian Estimation with prior Markov Random Field models. In this approach, the solution of a processing problem is characterized as the minimizer of a cost function which has two types of terms: terms that specify that the solution should be compatible with the available observations and terms that impose certain constraints on the solution. In this paper we show that by the appropriate choice of these terms, one can use this approach in almost every processing step for accurate interferogram demodulation. Specifically, one can construct: robust smoothing filters that are almost insensitive to edge effects; operators that automatically determine a mask that indicates the shape of the region where valid fringes are available; adaptive quadrature filters for phase recovery from single and multi-phase stepping interferograms and robust phase unwrapping algorithms. !12
展开▼
