Boundary detection through dynamic polygons

by Antonio Pievatolo (Milano, Italia) and Peter Green (Bristol, UK). Journal of the Royal Statistical Society, B, 60, 609-626 (1998).
A method for the Bayesian restoration of noisy binary images portraying an object with constant greylevel on a background is presented. The restoration, performed by fitting a polygon with any number of sides to the object outline, is driven by a new probabilistic model for the generation of polygons in a compact subset of R^2, which is used as a prior distribution for the polygon. The simulation from the prior and the calculation of the a posteriori mean of greylevels are carried out through reversible jump MCMC computation. Two examples of restoration of synthetic images are presented.
Back to Peter Green's research page