K. Wu, M.D. Levine Efficient object recognition usually requires an object description that captures only coarse information of object shape. In this research, we propose a new model called parametric geons as volumetric primitives to characterize the shape of object parts. Parametric geons are seven qualitative shape types regarded as basic volumetric forms in the art of sculpture. They are defined by constrained superellipsoids, with deformation parameters controlling tapering and bending. Model recovery is achieved by fitting all of the seven models to rangefinder data using a stochastic global optimization method (Very Fast Simulated Re-Annealing) and then selecting the best model according to the minimum fitting residual. The advantage of using parametric geons is that it provides a global shape constraint which protects model recovery from the influence of noise and minor shape variations and allows explicit model shape verification. Thus a volumetric shape can be robustly approximated by one of the seven distinct shape types. The resulting model can be used as a coarse description of object parts in the process of qualitative object recognition.