A non-iterative method for the estimation of superquadric parameters from depth maps

In this work we present a method for the extraction of superquadric shape and position parameters from depth maps. The method, which is non-iterative, makes use of the information obtained from the 3-D silhouette of the objects in order to estimate shape parameters and center position. We discuss the applicability conditions and propose a measure of the reliability of the obtained estimation. Some experimental results obtained from synthetically realized scenes are presented.