Meshless parameterization for dimensional reduction integrated in 3D voxel reconstruction using a single PC

Abstract

Shape-From-Silhouettes (SFS) is one of the most popular ideas for reconstructing the 3D voxel of an object from silhouettes images. This paper presents a method based SFS and pre-computing methods for 3D voxel reconstruction using a single PC. This method is reduced the memory usage. Towards this approach, a meshless parameterization for dimensional reduction is integrated in the process to obtain object representation in 2D form. Since the meshless parameterization requires the solution of large linear system on a whole 3D voxel, by taking the advantages of 3D voxel reconstruction process, the meshless parameterization is computed locally. The proposed method is able to implement an optimize system and utilize in various applications. © Springer-Verlag Berlin Heidelberg 2007.