Photo-Realistic Rendering of Fiber Assemblies

Dissertation, Universität Bonn, Jan. 2008
 

Abstract

In this thesis we introduce a novel uniform formalism for light scattering from filaments, the Bidirectional Fiber Scattering Distribution Function (BFSDF). Similar to the role of the Bidirectional Surface Scattering Reflectance Distribution Function (BSSRDF) for surfaces, the BFSDF can be seen as a general approach for describing light scattering from filaments. Based on this theoretical foundation, approximations for various levels of abstraction are derived allowing for efficient and accurate rendering of fiber assemblies, such as hair or fur. In this context novel rendering techniques accounting for all prominent effects of local and global illumination are presented. Moreover, physically-based analytical BFSDF models for human hair and other kinds of fibers are derived. Finally, using the model for human hair we make a first step towards image-based BFSDF reconstruction, where optical properties of a single strand are estimated from “synthetic photographs” (renderings) a full hairstyle.

Keywords: hair modelling, hair rendering

Download: http://nbn-resolving.de/urn:nbn:de:hbz:5N-13198

Bibtex

@PHDTHESIS{Zinke2008a,
    author = {Zinke, Arno},
     title = {Photo-Realistic Rendering of Fiber Assemblies},
      type = {Dissertation},
      year = {2008},
     month = jan,
    school = {Universit{\"a}t Bonn},
  keywords = {hair modelling, hair rendering},
  abstract = {In this thesis we introduce a novel uniform formalism for light scattering from filaments, the
              Bidirectional Fiber Scattering Distribution Function (BFSDF). Similar to the role of the
              Bidirectional Surface Scattering Reflectance Distribution Function (BSSRDF) for surfaces, the BFSDF
              can be seen as a general approach for describing light scattering from filaments. Based on this
              theoretical foundation, approximations for various levels of abstraction are derived allowing for
              efficient and accurate rendering of fiber assemblies, such as hair or fur. In this context novel
              rendering techniques accounting for all prominent effects of local and global illumination are
              presented. Moreover, physically-based analytical BFSDF models for human hair and other kinds of
              fibers are derived. Finally, using the model for human hair we make a first step towards image-based
              BFSDF reconstruction, where optical properties of a single strand are estimated from ``synthetic
              photographs'' (renderings) a full hairstyle.},
       url = {http://nbn-resolving.de/urn:nbn:de:hbz:5N-13198}
}