An Ellipsoid Shape Approximation Approach for 3d Shape Representation

Shape representation plays an important role to preserve shape geometry information of any object. The majority of existing shape representation methods used sphere to approximate 3D object. Sphere is divergence in nature and hence the mapping between sphere and object is noncompact and the resultant shape representation becomes less informative. In this paper, we propose a shape representation method based on ellipsoid to approximate 3D object. Ellipsoid is convergence in nature and it is better fitted to the body of the shape rather than sphere. The proposed shape representation method using ellipsoid is able to generate informative shape sampling function capturing more detail information smoothly of an object. Our proposed method decreases the approximation error than the conventional spherical based method.