Fran¸cois Sillion, James Arvo, Stephen Westin, and Donald P. Greenberg. A global illumination solution for general reflectance distributions. Computer Graphics, 25(4):187--196, July 1991.  Home/Search   Document Not in Database   Summary   Related Articles   Check

....finely subdivided. Work needs to be done to optimize these algorithms for complex scenes (m on the order of hundreds of thousands or millions of polygons) The radiosity method has been generalized to non diffuse scenes using spherical harmonics as an approximation to directional distributions [Sillion et al. 91] Higher degree elements [Heckbert Winget91,Max Allison92,Salesin et al. 92,Troutman Max93] Galerkin methods [Heckbert Winget91,Zatz93] and improved integration techniques [Schroder93,Pietrek93] have been used to improve the accuracy of an approximation. The following two chapters discuss ....

Fran¸cois X. Sillion, James R. Arvo, Stephen H. Westin, and Donald P. Greenberg. A global illumination solution for general reflectance distributions. Computer Graphics (SIGGRAPH '91 Proceedings), 25(4):187--196, July 1991.

....Though more expensive, the simulation of directional reflection is essential for realistic image synthesis. One promising approach to solving directional light transport is the finite element method, as pioneered by Immel et al. 16] and Shao et al. 21] and later refined by Sillion et al. [22]. Recently, Gortler et al. 14] and Schroder et al. 20] proposed an algorithm based on wavelets that focuses effort on the significant energy transfers, for the simpler case of radiosity. These works use the nonstandard decomposition of the transport operator, and represent radiosity as a ....

....polynomials, and Gortler et al. 14] used wavelets. In the more general context of radiance, the distribution of light leaving a patch has both spatial and angular variation. Immel et al. 16] used piecewise constant basis functions for both spatial and angular variation. Later, Sillion et al. [22] used spherical harmonics for the angular variation and piecewise constant basis functions for the spatial variation. In Sect. 3 we motivate and introduce our choice of basis, a wavelet basis for both spatial and angular variation. Regardless of the choice of basis functions, we can obtain a ....

[Article contains additional citation context not shown here]

Fran¸cois X. Sillion, James R. Arvo, Stephen H. Westin, and Donald P. Greenberg. A global illumination solution for general reflectance distributions. In Proceedings of SIGGRAPH'91, pages 187--196, July 1991.

....is O(k 2 v) although so far they have only been applied to the Haar basis. Related methods partition the scene to improve efficiency [57] There are other generalizations of the radiosity method that expand its capabilities beyond diffuse surfaces to surfaces with general reflectance [49], and to simulation of the volumetric scattering, absorption, and emission of light in participating media [39] but we have ignored these techniques because they are very different from and much more complex than diffuse surface radiosity algorithms. It is also possible to simulate diffuse ....

François X. Sillion, James R. Arvo, Stephen H. Westin, and Donald P. Greenberg. A global illumination solution for general reflectance distributions. Computer Graphics (SIGGRAPH '91 Proceedings), 25(4):187--196, July 1991.

....summarized below. The result is an efficient algorithm, which we demonstrate using both simple and complex scenes. The first choice is in the parameterization of the unknown light distribution. One possibility is to use radiance distributions, which are functions of surface position and direction [23, 34]. The alternative is to use two point transport intensities, which are functions of two surface positions [2, 32] We describe our motivation for using radiance distributions. A second area of choice in designing a glossy global illumination algorithm is that of basis functions. One can use a ....

....can be discretized to facilitate representation and efficient transport. 2.1 Radiance distributions vs. two point transport intensities Two fundamentally different representations of the transported light have been used for glossy global illumination. Immel et al. 23] and Sillion et al. [34] represent the light in the scene as radiance distributions, a function of two spatial and two angular variables on each surface patch. By contrast, Aupperle and Hanrahan [2] Pattanaik and Bouatouch [28] and Schroder and Hanrahan [32] use a two point transport intensity, a function of four ....

[Article contains additional citation context not shown here]

Fran¸cois X. Sillion, James R. Arvo, Stephen H. Westin, and Donald P. Greenberg. A global illumination solution for general reflectance distributions. In Proceedings of SIGGRAPH 91, pages 187--196, July 1991.

....[13, 38] feature nearly mutually exclusive models of light reflection: each valid in itself, but incomplete. Methods that combine ray tracing and radiosity typically neglect more complex modes of reflection [53] The rendering equation [19] and methods that incorporate complex surface reflection [4, 52] still neglect light scattering by participating media such as smoke. Methods that model participating media have thus far limited the forms of surface reflection [46, 47, 21] Finally, absent from all of the cited approaches are macroscopic effects due to interference and diffraction. Every ....

....effects such as local scattering and absorption, which then enter into the simulations at the macroscopic scale. For instance, the physically based reflection model of He [16] employs wave optics to characterize reflection from rough surfaces, and this model can be used for global illumination [52]. For many materials with complex microgeometries, there exists a natural hierarchy of scales, with geometrical optics taking over at the point when wave effects become negligible [60] Another avenue by which physical optics effects can enter into radiative transfer is through physical ....

Fran¸cois Sillion, James Arvo, Stephen Westin, and Donald P. Greenberg. A global illumination solution for general reflectance distributions. Computer Graphics, 25(4):187--196, July 1991.

....algorithms for hierarchical solutions [10, 6] avoid the quadratic cost by first clustering the environment and then refining the clusters. Nonetheless, little work has been performed for non diffuse environments. Twopass algorithms [9, 11] and a general solution using directional representations [7] have treated more general environments in the context of progressive refinement radiosity. A hierarchical solution to general environments has also been proposed [1] but in the case of that algorithm the initial linking cost becomes O(n 3 ) in the number of initial polygons, making it unusable ....

....extension of previous clustering methods since, as noted before [6] clusters do not behave as isotropic scatterers, even if composed solely of diffuse surfaces. It is based on the representation of radiant intensity by directional distribution functions, and extends the spirit presented in [7] to hierarchical clustering. The result is the first efficient hierarchical algorithm permitting the efficient of complex, non diffuse environments. In addition, this representation affords a smooth transition between the representation at the level of (non diffuse and diffuse) surfaces to the # ....

[Article contains additional citation context not shown here]

François Sillion, James Arvo, Stephen Westin, and Donald P. Greenberg. A global illumination solution for general reflectance distributions. Computer Graphics, 25(4):187--196, August 1991. Proceedings SIGGRAPH '91 in Las Vegas (USA).

Online articles have much greater impact   More about CiteSeer.IST   Add search form to your site   Submit documents   Feedback  

CiteSeer.IST - Copyright Penn State and NEC