## Towards Interactive Photorealistic Rendering of Indoor Scenes: A Hybrid Approach (1999)

### Cached

### Download Links

- [www.sci.utah.edu]
- [www.cs.utah.edu]
- [pdf.aminer.org]
- [www.cs.utah.edu]
- DBLP

### Other Repositories/Bibliography

Citations: | 23 - 1 self |

### BibTeX

@INPROCEEDINGS{Udeshi99towardsinteractive,

author = {Tushar Udeshi and Charles D. Hansen},

title = {Towards Interactive Photorealistic Rendering of Indoor Scenes: A Hybrid Approach},

booktitle = {},

year = {1999},

pages = {63--76}

}

### OpenURL

### Abstract

. Photorealistic rendering methods produce accurate solutions to the rendering equation but are computationally expensive and typically noninteractive. Some researchers have used graphics hardware to obtain photorealistic effects but not at interactive frame rates. We describe a technique to achieve near photorealism of simple indoor scenes at interactive rates using both CPUs and graphics hardware in parallel. This allows the user the ability to interactively move objects and lights in the scene. Our goal is to introduce as many global illumination effects as possible while maintaining a high frame rate. We describe methods to generate soft shadows, approximate one-bounce indirect lighting, and specular reflection and refraction effects. 1 Introduction and previous work Research in photorealistic rendering has concentrated on numerically solving the rendering equation[9]. Ray tracers[22] use Monte Carlo methods while radiosity systems[6] use finite element methods. These give accurat...

### Citations

752 | The rendering equation
- Kajiya
(Show Context)
Citation Context ... indirect lighting, and specular reflection and refraction effects. 1 Introduction and previous work Research in photorealistic rendering has concentrated on numerically solving the rendering equation=-=[9]-=-. Ray tracers[22] use Monte Carlo methods while radiosity systems[6] use finite element methods. These give accurate solutions but are computationally expensive. Classic radiosity methods typically co... |

586 |
An Improved Illumination Model for Shaded Display
- Whitted
- 1980
(Show Context)
Citation Context ...ng, and specular reflection and refraction effects. 1 Introduction and previous work Research in photorealistic rendering has concentrated on numerically solving the rendering equation[9]. Ray tracers=-=[22]-=- use Monte Carlo methods while radiosity systems[6] use finite element methods. These give accurate solutions but are computationally expensive. Classic radiosity methods typically converge faster to ... |

350 | Modeling the interaction of light between diffuse surfaces
- Goral, Torrance, et al.
- 1984
(Show Context)
Citation Context ...1 Introduction and previous work Research in photorealistic rendering has concentrated on numerically solving the rendering equation[9]. Ray tracers[22] use Monte Carlo methods while radiosity systems=-=[6]-=- use finite element methods. These give accurate solutions but are computationally expensive. Classic radiosity methods typically converge faster to a solution to the rendering equation than ray trace... |

346 | Casting curved shadows on curved surfaces
- WILLIAMS
- 1978
(Show Context)
Citation Context ...s unsuitable for scenes with moving objects. Shadow map / Shadow buffer algorithm. Shadow maps use the depth buffer and projective texture mapping to create a screen space method for shadowing objects=-=[15, 16, 23]-=-. The scene is rendered from a light source L and also from the eye. The depth values of pixels from the eye are transformed to the light-view coordinate system and then compared to depth values recor... |

257 | algorithms for computer graphics
- Crow, Shadow
- 1977
(Show Context)
Citation Context ...en if one object in the scene moves, the entire shadow map needs to be recalculated. Therefore, this algorithm is not suitable for dynamic scenes. Shadow volume algorithm. The shadow volume algorithm =-=[4] is a-=-n object-space technique for generating shadows. A shadow volume, defined with respect to an occludersO and a point light source L, is that region in space where L is blocked by O. 2 Fig. 1. "Sha... |

238 | Towards Image Realism with Interactive Update Rates in Complex Virtual Building Environments
- Airey, Rohlf, et al.
- 1990
(Show Context)
Citation Context ...ese limitations but require path tracing[9] to get indirect illumination effects which introduces noise and is computationally expensive. Many interactive environments such as Virtual Building Systems=-=[1]-=- rely on precomputation of static environments to form progressive radiosity solutions. These suffer from large computational overhead and unchangeable geometry. Even in incremental radiosity solution... |

215 |
P.: Fast Shadows and Lighting Effects using Texture Mapping
- Segal, Korobkin, et al.
- 1992
(Show Context)
Citation Context ...is unsuitable for scenes with moving objects. Shadow map/ Shadow buffer algorithm. Shadow maps use the depth buffer and projective texture mapping to create a screen space method for shadowing objects=-=[13, 14]-=-. The scene is rendered from a light source(L) and also from the eye. The depth values of pixels corresponding to the same region in object space are compared. If the depth value from the eye is less,... |

209 |
Rendering antialiased shadows with depth maps
- Reeves, Salesin, et al.
- 1987
(Show Context)
Citation Context ...s unsuitable for scenes with moving objects. Shadow map / Shadow buffer algorithm. Shadow maps use the depth buffer and projective texture mapping to create a screen space method for shadowing objects=-=[15, 16, 23]-=-. The scene is rendered from a light source L and also from the eye. The depth values of pixels from the eye are transformed to the light-view coordinate system and then compared to depth values recor... |

189 | Instant radiosity
- Keller
- 1997
(Show Context)
Citation Context ..., precomputed radiosity solution for indirect lighting. While this is acceptable for walk-throughs and minor scene geometry changes, it is unsuitable for scenes in which light sources can move. Keller=-=[10]-=- suggested a method to get fast solutions for indirect illumination by using hardware-assisted particle tracing. Particles are shot from the light in software. When a particle hits a diffuse object in... |

170 | Tracing Polygonal Objects
- Heckbert, Hanrahan, et al.
(Show Context)
Citation Context ...ge. This gives a good approximation for reflective surfaces. However for refraction, this approach is an acceptable approximation only if the rays incident on to the refractive surface are para-axial =-=[7]-=-, which requires all refractive surfaces to be perpendicular to the line of sight vector. This approach cannot be used for non-planar surfaces. Moreover each reflective/refractive surface visible from... |

161 | Interactive ray tracing
- Parker, Martin, et al.
- 1999
(Show Context)
Citation Context ... require significant recomputation time. Moreover these solutions do not account for specular reflection and refraction. Systems based on ray tracing are mostly non-interactive. Recently Parker et al.=-=[14]-=- developed an interactive ray tracer. However it does not account for indirect illumination. Moreover, moving objects must be spatially bounded which implies that all motion is known a priori. Hardwar... |

134 | Some techniques for shading machine renderings of solids
- Appel
- 1968
(Show Context)
Citation Context ...solution to these is given in Section 2. We use multiple pipes and CPUs to implement both the shadow map and the shadow volume algorithm. The details are in Section 4.1. 1.2 Indirect lighting As Appel=-=[2]-=- recognized, greater realism requires global illumination models, which account for inter-reflection of light between surfaces. Studies[19] have shown that indirect illumination also gives an importan... |

125 | Parallel volume rendering using binary-swap compositing
- Ma, Painter, et al.
- 1994
(Show Context)
Citation Context ...splay pipe. The partial images rendered in the pipes are read into shared memory. The composition of these involves just an additive blending. We implemented this in two ways: hierarchical binary swap=-=[11, 20]-=- which used graphics hardware and software compositing. When enough CPUs are available, the latter results in a higher frame rate. This is because a few CPUs can be assigned to the task of compositing... |

108 | Improved computational methods for ray tracing
- Weghorst, Hooper, et al.
- 1984
(Show Context)
Citation Context ... which is received from dl and reflected by P . We use a single-plane algorithm[17] to calculate the form-factors. Each polygon is assigned a weight initialized to zero. A wide-perspective item buffer=-=[21]-=- is rendered with dl as the center of projection. The form-factor from dl to each pixel in the item buffer is precomputed in a map with the help of Nusselt Analog. The weight of a polygon P is the the... |

75 | Incremental Radiosity: An Extension of Progressive Radiosity to an Interactive Image
- Chen
- 1990
(Show Context)
Citation Context ...rely on precomputation of static environments to form progressive radiosity solutions. These suffer from large computational overhead and unchangeable geometry. Even in incremental radiosity solutions=-=[3]-=-, geometry changes require significant recomputation time. Moreover these solutions do not account for specular reflection and refraction. Systems based on ray tracing are mostly non-interactive. Rece... |

66 | Pipeline Rendering: Interaction and Realism Through Hardware-based Multi-Pass perspective view orthographic view
- Diefenbach
(Show Context)
Citation Context ...which implies that all motion is known a priori. Hardware-based solutions, which are interactive, are inherently of lower quality due to the limited feature set of the graphics accelerator. Diefenbach=-=[5]-=- demonstrated techniques of using graphics hardware to get specular reflection, refraction, caustics and transparency but not at interactive frame rates. In this paper, we describe techniques of using... |

30 |
Half-tone representation of 3-d objects illuminated by area sources or polyhedron sources
- Nishita, Nakamae
- 1983
(Show Context)
Citation Context ... into three regions: totally occluded, partially occluded and totally unoccluded. This can be accomplished by rendering those shadow volumes which form umbral or penumbral boundaries, as described in =-=[12]-=-. The number of shadow volumes would be significantly reduced. However a fast way to render the partially occluded regions of the scene is still an open question. Acknowledgments This work was support... |

24 |
shadows, real time
- Real
- 1991
(Show Context)
Citation Context ...ing shadow quads which are closer to the eye than any object in the scene. We refer to this as s-count. If the s-count of a particular pixel is positive, we infer that the pixel is in shadow. Heidmann=-=[18]-=- used the stencil buffer to determine s-count for each pixel. In our implementation we use the red channel in the back buffer to maintain this count. The scene is rendered in the front buffer and the ... |

17 |
Sillion and Claude Puech. A general two-pass method integrating specular and diffuse reflection
- François
- 1989
(Show Context)
Citation Context ...o P . Therefore, this form-factor multiplied by the reflectivity of P and the emissivity of dl is a measure of the energy which is received from dl and reflected by P . We use a single-plane algorithm=-=[17]-=- to calculate the form-factors. Each polygon is assigned a weight initialized to zero. A wide-perspective item buffer[21] is rendered with dl as the center of projection. The form-factor from dl to ea... |

13 | Fast Soft Shadows
- Herf, Heckbert
(Show Context)
Citation Context ...w buffer and shadow volume. We give a brief description of these algorithms in the following paragraphs. Projective textures. This algorithm has been used to generate soft shadows during walk-throughs=-=[8]-=-. To find the shadows cast on a particular polygon P , with respect to a light source L, all other polygons in the scene are projected on to P , with L as the center of projection. This can be done in... |

10 |
Parallel Multipipe Rendering for Very Large Isosurface Visualization
- Udeshi, Hansen
- 1999
(Show Context)
Citation Context ...splay pipe. The partial images rendered in the pipes are read into shared memory. The composition of these involves just an additive blending. We implemented this in two ways: hierarchical binary swap=-=[11, 20]-=- which used graphics hardware and software compositing. When enough CPUs are available, the latter results in a higher frame rate. This is because a few CPUs can be assigned to the task of compositing... |

9 | Visual glue
- Thompson, Shirley, et al.
- 1998
(Show Context)
Citation Context ...The details are in Section 4.1. 1.2 Indirect lighting As Appel[2] recognized, greater realism requires global illumination models, which account for inter-reflection of light between surfaces. Studies=-=[19]-=- have shown that indirect illumination also gives an important visual cue signaling contact between two surfaces. Radiosity systems give accurate solutions for indirect lighting in diffuse scenes but ... |

1 |
Pipeline rendering: Interaction and realism through hardware-based multi-pass rendering
- Deifenbach
- 1996
(Show Context)
Citation Context ...which implies that all motion is known a priori. Hardware-based solutions, which are interactive, are inherently of lower quality due to the limited feature set of the graphics accelerator. Diefenbach=-=[5]-=- demonstrated techniques of using graphics hardware to get specular reflection, refraction , caustics and transparency but not at interactive frame rates. In this paper, we describe techniques of usin... |