Conference Paper

Interactive Rendering of Translucent Objects.

Max-Planck-Inst. fur Inf., Saarbrucken, Germany
DOI: 10.1109/PCCGA.2002.1167862 Conference: 10th Pacific Conference on Computer Graphics and Applications (PG 2002), 9-11 October 2002, Beijing, China
Source: DBLP

ABSTRACT This paper presents a rendering method for translucent objects, in which view point and illumination can be modified at interactive rates. In a preprocessing step the impulse response to incoming light impinging at each surface point is computed and stored in two different ways: The local effect on close-by surface points is modeled as a per-texel filter kernel that is applied to a texture map representing the incident illumination. The global response (i.e. light shining through the object) is stored as vertex-to-vertex throughput factors for the triangle mesh of the object. During rendering, the illumination map for the object is computed according to the current lighting situation and then filtered by the precomputed kernels. The illumination map is also used to derive the incident illumination on the vertices which is distributed via the vertex-to-vertex throughput factors to the other vertices. The final image is obtained by combining the local and global response. We demonstrate the performance of our method for several models.

  • [Show abstract] [Hide abstract]
    ABSTRACT: Light propagation in scenes with translucent objects is hard to model efficiently for interactive applications. The inter-reflections between objects and their environments and the subsurface scattering through the materials intertwine to produce visual effects like color bleeding, light glows and soft shading. Monte-Carlo based approaches have demonstrated impressive results but are computationally expensive, and faster approaches model either only inter-reflections or only subsurface scattering. In this paper, we present a simple analytic model that combines diffuse inter-reflections and isotropic subsurface scattering. Our approach extends the classical work in radiosity by including a subsurface scattering matrix that operates in conjunction with the traditional form-factor matrix. This subsurface scattering matrix can be constructed using analytic, measurement-based or simulation-based models and can capture both homogeneous and heterogeneous translucencies. Using a fast iterative solution to radiosity, we demonstrate scene relighting and dynamically varying object translucencies at near interactive rates.
    Proceedings of the ACM SIGGRAPH Symposium on Interactive 3D Graphics and Games; 03/2013
  • Source
    [Show abstract] [Hide abstract]
    ABSTRACT: Subsurface scattering is a complex physical process, which, in many cases, significantly affects the appearance of certain materials. In the pursuit of generating ever more realistic scenes, it is a phenomenon which must be incorporated into rendering frameworks. However, the complexity of the physical process which causes subsurface scattering has a tremendous effect on rendering time. Though such rendering costs are acceptable in many settings, they are far to high for any process which requires interactive visualization. In this paper, we propose a very simple approximation of subsurface scattering which requires a pre-computation step but incurs absolutely no runtime overhead. We illustrate the basic model used to derive our results, as well as a spectral mesh processing framework which generalizes our model while producing images of higher visual fidelity. This makes it possible to incorporate sub- surface scattering effects into real-time visualizations where performance is the primary goal. We hypothesize that the complexity of the underlying physical process of subsurface scattering results in both highly apparent and subtle effects. Accepting the assumption that the realism of our subsurface scattering effect is secondary in importance to the speed at which it can be displayed, we make sacrifices in the physical realism of our approximation for the sake of speed. The results, however, maintain the overall landmark effects of subsurface scattering and provide the ability to produce visually compelling results with no runtime overhead. There is a fundamental dichotomy in computer graphics between phys- ical accuracy and interactivity. The greater the degree to which reality is approximated in the process of rendering, the more computation that rendering necessarily requires. Even the generality of the ren- dering equation makes the simplication of geometric optics, sacric- ing some expressibility for the sake of simplicity. The ways in which this dichotomy has most frequently been addressed are through pre- processing and approximation. Sometimes, it is possible to capture complex physically based rendering effects through a pre-processing phase; the results of which can be evaluated, displayed, and possibly modied in real-time. For example, precomputed radiance transfer allows for complex lighting effects to be evaluated at runtime, assum- ing that the geometry and lighting environment, and the manner in which they may change, are known a priori. Thus, realistic lighting effects which would be impossible to compute at runtime can still be displayed at interactive rates. Another facet of precomputed radiance transfer, however, is that of approximation. In order to allow the pre- computed results to be evaluated and displayed in real-time, an ap- propriate representation for these results, such as spherical harmonics, must be utilized. The spherical harmonics basis allows for the efcient storage and evaluation of many precomputed radiance transfer results. However, this efcenc y is achieved at the cost of its low-frequency ap- proximation of the original lighting signal. The decision as to which methods to use to simulate complex physical effects in the process of rendering is often reduced to an application dependent cost-benet analysis, where the designer of the application must weigh the level of accuracy necessary against the level of interactivity required for the application. One such computationally expensive rendering effect is subsurface scattering. Subsurface scattering is a physical property of the interac- tion of light with different materials. It occurs when light is incident upon a material with some degree of translucency. Instead of strictly reecting off the surface of the material, some amount of light will ac- tually enter the material, scattering and diffusing as it moves, and will then exit at points distinct from where it entered. Subsurface scattering produces visual effects which are signicant enough to warrent imple- mentation in many different visualization applications. When physical accuracy and realism are the foremost goals, many high delity ap- proximations to subsurface scattering may be used to achieve highly convincing results, such as (3). However, even the fastest existing methods of approximating subsurface scattering result in a decrease in rendering speeds by a factor of two or more. Main Results: In this paper, we present an approximation to sub- surface scattering which allows results generated in a pre-processing phase to be expressed using the standard local illumination model (1). Our method computes the effects of subsurface scattering at a given surface point by incorporating information about the incident light at neighboring surface points. We present both a nav e implemen- tation, using a weighted sum of neighboring normals, and a more rened, spectral mesh processing implementation, which produces higher quality results by eliminating certain descretization artifacts. Further, our spectral processing approach allows for the ltering ker- nel being used to create the subsurface scattering effect to be changed at a near interactive rate. Both approaches produce a set of ltered normals, which need only be used in place of the original normals of the mesh to display the subsurface scattering effect. The fact that the effect is achieved by using these pre-ltered normals during the rendering phase, means that no overhead, beyond that of the standard local illumination model, is incurred. Thus, runtime framerates are not at all affected, and a coarse, but ìfreeî, approximation of subsurface scattering is provided.
  • Source

Full-text (2 Sources)

Available from
May 21, 2014