“Implementation of Modeling Hair from Multiple Views” by Wei, Ofek, Quan and Shum
Conference:
Title:
- Implementation of Modeling Hair from Multiple Views
Session/Category Title: Image-Based Rendering
Presenter(s)/Author(s):
Interest Area:
- Implementation Sketch
Abstract:
Modeling hair is difficult due to its complex geometry and re- flectance properties [Marschner et al. 2003]. It often requires in- tensive user interactions. Paris et al. [Paris et al. 2004] developed an approach to capture hair geometry from a fixed viewpoint under controlled variable illumination. We developed a more practical approach in [Wei et al. 2005] that uses a hand-held camera under natural lighting conditions and does not require a specific setup. The success of this approach depends on good solutions for the following issues.
References:
Hartley, R., and Zisserman, A. 2000. Multiple View Geometry in Computer Vision. Cambridge University Press.
Lhuillier, M., and Quan, L. 2005. A quasi-dense approach to surface reconstruction from uncalibrated images. IEEE Trans, on Pattern Analysis and Machine Intelligence vol 27, 3, pages 418–433.
Marschner, S., Jensen, H., cammarano, M., Worley, S., and Hanrahan, P. 2003. Light scattering from human hair fibers. ACM Trans. on Graphics 3, 22, 780–791.
Paris, S., Briceno, H. M., and Sillion, F. X. 2004. Capture of hair geometry from multiple images. In Proc. of SIGGRAPH conference, ACM, 712–719.
Wei, Y., Ofek, E., Quan, L., and Shum, H. 2005. Modeling hair from multiple views. In Proc. of SIGGRAPH conference, ACM.
