Article ID: | iaor2006878 |
Country: | Cuba |
Volume: | 26 |
Issue: | 1 |
Start Page Number: | 79 |
End Page Number: | 89 |
Publication Date: | Jan 2005 |
Journal: | Revista de Investigacin Operacional |
Authors: | Amargs Joaquin D. Pina |
Keywords: | artificial intelligence: expert systems |
A new method to solve the branching problem in surfaces of three-dimensional (3D) models reconstructed from parallel cross sections is presented in this paper. This is an important problem in medical imaging and other object-modelling applications, where the reconstruction of the surface of 3D models from images that represent a set of parallel planes is of interest. The proposed method is based on the skeletonization technique to create new contours, corresponding to an artificial intermediate slice that models the level where branching occurs. Some experimental results of the application of the proposed method to solve the branching problem on synthetic complex examples and actual medical imaging data are shown.