NONLINEAR GENERALIZATION OF POINT DISTRIBUTION MODELS USING POLYNOMIAL REGRESSION

We have previously described how to model shape variability by means o f point distribution models (PDM) in which there is a linear relations hip between a set of shape parameters and the positions of points on t he shape. This linear formulation can fail for shapes which articulate or bend. We show examples of such failure for both real and synthetic classes of shape. A new, more general formulation for PDMs, based on polynomial regression, is presented. The resulting polynomial regressi on PDMs (PRPDM) perform well on the data for which the linear method f ailed.