3-D Scene Modelling from Multiple Images using Radial Basis Function Networks

A new approach for modelling multiple 3-D objects from a sparse set of images taken from various viewpoints is proposed in this paper. A voxel model of the scene is estimated from the given set of images using the space carving algorithm. An implicit radial basis function (RBF) network is used afterwards to model the voxel data. The multiorder function is chosen as the kernel function due to its property of enforcing smoothing constraints in the first three derivatives. A suitable initialization is proposed for the RBF parameters. Displacements of surface patches along epipolar lines are used to update the centers of basis functions leading to the modelling errors minimization. The proposed method is used to model a complex 3-D scene.