“Surface network construction from non-parallel cross-sections” by Liu, Ju, Low and Bajaj

  • ©Lu Liu, Tao Ju, Daniel Low, and Chandrajit Bajaj

Conference:


Type(s):


Title:

    Surface network construction from non-parallel cross-sections

Presenter(s)/Author(s):



Abstract:


    Surface construction from incomplete data is a topic of wide interest in geometry processing. A common class of incomplete inputs that arises in several application domains, including bio-medicine and geology, is a stack of curves representing the planar cross-sections of a complete surface. In medical imaging, for example, these curves are typically hand-drawn by physicians to delineate contours of anatomical structures on 2D cross-sections of a 3D image volume generated by MRI, CT or ultrasound. Given a stack of planar cross-sections, a complete 3D surface is desired that connects the curves on each cross-section plane. In particular, such surface needs to be both topologically correct (i.e., closed and interpolating input curves) and geometrically smooth.

References:


    1. Barequet, G., Goodrich, M. T., Levi-Steiner, A., and Steiner, D. 2003. Straight-skeleton based contour interpolation. In SODA ’03: Proceedings of the fourteenth annual ACM-SIAM symposium on Discrete algorithms, 119–127.
    2. Ju, T., Warren, J. D., Carson, J., Eichele, G., Thaller, C., Chiu, W., Bello, M., and Kakadiaris, I. A. 2005. Building 3d surface networks from 2d curve networks with application to anatomical modeling. The Visual Computer 21, 8–10, 764–773.


ACM Digital Library Publication:



Overview Page: