Game Development Reference
In-Depth Information
implemented independently, but they obviously could be implemented in a single
pass as well. Moreover, the gradient-directed shock filter provides finer artistic
control. The parameter σ g restricts smoothing to the major eigenvector direc-
tion. This is especially useful for preserving small image features. To achieve a
stronger abstraction, the isotropic smoothing parameter σ i is useful.
5.5
Conclusion
In this chapter, an automatic technique for image and video abstraction, based
on adaptively controlled flow-guided smoothing and directional shock filtering,
was presented. It aggressively smoothes out unimportant image regions, but
it protects important features by enhancing contrast and directional coherence,
providing a good balance between content abstraction and feature enhancement
consistently across the image. For abstraction at the level of the anisotropic
Kuwahara filter, the GPU implementation processes video in real time and creates
temporally coherent output without further processing.
5.6
Acknowledgments
Original photographs from flickr.com kindly provided under Creative Commons
license by Tambako the Jaguar ( Figure 5.2(a) ) and Ivan Mlinaric ( Figure 5.10 ) .
Original photograph in Figure 5.2(b) courtesy of Phillip Greenspun.
Bibliography
[Briggs et al. 00] W. L. Briggs, V. E. Henson, and S. F. McCormick. “A Multigrid
Tutorial.” Presentation,
SIAM Annual Meeting, Rio Grande, Puerto Rio,
July 2000.
[Brox et al. 06] T. Brox, R. van den Boomgaard, F. Lauze, J. van de Weijer,
J.Weickert,P.Mrazek, and P. Kornprobst. “Adaptive Structure Tensors
and Their Applications.” In Visualization and Processing of Tensor Fields ,
edited by Joachim Weickert and Hans Hagen, pp. 17-47. Berlin: Springer,
2006.
[Cabral and Leedom 93] B. Cabral and L. C. Leedom. “Imaging Vector Fields
Using Line Integral Convolution.” In Proceedings of the 20th Annual Confer-
ence on Computer Graphics and Interactive Techniques , pp. 263-270. New
York: ACM, 1993.
[Goldberg 91] D. Goldberg. “What Every Computer Scientist Should Know
about Floating-Point Arithmetic.”
ACM Computing Surveys 23:1 (1991),
5-48.
 
Search Nedrilad ::




Custom Search