Donald P. Greenberg

A progressive refinement approach to fast radiosity image generation

A reformulated radiosity algorithm is presented that produces initial images in time linear to the number of patches. The enormous memory costs of the radiosity algorithm are also eliminated by computing form-factors on-the-fly. The technique is based on the approach of rendering by progressive refinement. The algorithm provides a useful solution almost immediately which progresses gracefully and continuously to the complete radiosity solution. In this way the competing demands of realism and interactivity are accommodated. The technique brings the use of radiosity for interactive rendering within reach and has implications for the use and development of current and future graphics workstations.

A comprehensive physical model for light reflection

A new general reflectance model for computer graphics is presented. The model is based on physical optics and describes specular, directional diffuse, and uniform diffuse reflection by a surface. The reflected light pattern depends on wavelength, incidence angle, two surface roughness parameters, and surface refractive index. The formulation is self consistent in terms of polarization, surface roughness, masking/shadowing, and energy. The model applies to a wide range of materials and surface finishes and provides a smooth transition from diffuse-like to specular reflection as the wavelength and incidence angle are increased or the surface roughness is decreased. The model is analytic and suitable for Computer Graphics applications. Predicted reflectance distributions compare favorably with experiment. The model is applied to metallic, nonmetallic, and plastic materials, with smooth and rough surfaces.

Non-linear approximation of reflectance functions

We introduce a new class of primitive functions with non-linear parameters for representing light reflectance functions. The functions are reciprocal, energy-conserving and expressive. They can capture important phenomena such as off-specular reflection, increasing reflectance and retro-reflection. We demonstrate this by fitting sums of primitive functions to a physically-based model and to actual measurements. The resulting representation is simple, compact and uniform. It can be applied efficiently in analytical and Monte Carlo computations. CR Categories: I.3.7 [Computer Graphics]: Three-Dimensional Graphics and Realism; I.3.3 [Computer Graphics]: Picture/Image Generation

A model of visual adaptation for realistic image synthesis

In this paper we develop a computational model of visual adaptation for realistic image synthesis based on psychophysical experiments. The model captures the changes in threshold visibility, color appearance, visual acuity, and sensitivity over time that are caused by the visual system’s adaptation mechanisms. We use the model to display the results of global illumination simulations illuminated at intensities ranging from daylight down to starlight. The resulting images better capture the visual characteristics of scenes viewed over a wide range of illumination levels. Because the model is based on psychophysical data it can be used to predict the visibility and appearance of scene features. This allows the model to be used as the basis of perceptually-based error metrics for limiting the precision of global illumination computations. CR

A multiscale model of adaptation and spatial vision for realistic image display

In this paper we develop a computational model of adaptation and spatial vision for realistic tone reproduction. The model is based on a multiscale representation of pattern, luminance, and color processing in the human visual system. We incorporate the model into a tone reproduction operator that maps the vast ranges of radiances found in real and synthetic scenes into the small fixed ranges available on conventional display devices such as CRT’s and printers. The model allows the operator to address the two major problems in realistic tone reproduction: wide absolute range and high dynamic range scenes can be displayed; and the displayed images match our perceptions of the scenes at both threshold and suprathreshold levels to the degree possible given a particular display device. Although in this paper we apply our visual model to the tone reproduction problem, the model is general and can be usefully applied to image quality metrics, image compression methods, and perceptually-based image synthesis algorithms. CR Categories: I.3.0 [Computer Graphics]: General;

Image-based BRDF measurement including human skin

We present a new image-based process for measuring the bidirectional reflectance of homogeneous surfaces rapidly, completely, and accurately. For simple sample shapes (spheres and cylinders) the method requires only a digital camera and a stable light source. Adding a 3D scanner allows a wide class of curved near-convex objects to be measured. With measurements for a variety of materials from paints to human skin, we demonstrate the new methods ability to achieve high resolution and accuracy over a large domain of illumination and reflection directions. We verify our measurements by tests of internal consistency and by comparison against measurements made using a gonioreflectometer.

Real-time robot motion planning using rasterizing computer graphics hardware

We present a real-time robot motion planner that is fast and complete to a resolution. The technique is guaranteed to find a path if one exists at the resolution, and all paths returned are safe. The planner can handle any polyhedral geometry of robot and obstacles, including disjoint and highly concave unions of polyhedra.The planner uses standard graphics hardware to rasterize configuration space obstacles into a series of bitmap slices, and then uses dynamic programming to create a navigation function (a discrete vector-valued function) and to calculate paths in this rasterized space. The motion paths which the planner produces are minimal with respect to an L1 (Manhattan) distance metric that includes rotation as well as translation.Several examples are shown illustrating the competence of the planner at generating planar rotational and translational plans for complex two and three dimensional robots. Dynamic motion sequences, including complicated and non-obvious backtracking solutions, can be executed in real time.

Spatiotemporal sensitivity and visual attention for efficient rendering of dynamic environments

We present a method to accelerate global illumination computation in prerendered animations by taking advantage of limitations of the human visual system. A spatiotemporal error tolerance map, constructed from psychophysical data based on velocity dependent contrast sensitivity, is used to accelerate rendering. The error map is augmented by a model of visual attention in order to account for the tracking behavior of the eye. Perceptual acceleration combined with good sampling protocols provide a global illumination solution feasible for use in animation. Results indicate an order of magnitude improvement in computational speed.

Perceiving spatial relationships in computer-generated images

The sources of visual information that must be present to correctly interpret spatial relations in images, the relative importance of different visual information sources with regard to metric judgments of spatial relations in images, and the ways that the task in which the images are used affect the visual informations usefulness are discussed. Cue theory, which states that the visual system computes the distances of objects in the environment based on information from the posture of the eyes and from the patterns of light projected onto the retinas by the environment, is presented. Three experiments in which the influence of pictorial cues on perceived spatial relations in computer-generated images was assessed are discussed. Each experiment examined the accuracy with which subjects matched the position, orientation, and size of a test object with a standard by interactively translating, rotating, and scaling the test object.<<ETX>>

A perceptually based physical error metric for realistic image synthesis

We introduce a new concept for accelerating realistic image synthesis algorithms. At the core of this procedure is a novel physical error metric that correctly predicts the perceptual threshold for detecting artifacts in scene features. Built into this metric is a computational model of the human visual systems loss of sensitivity at high background illumination levels, high spatial frequencies, and high contrast levels (visual masking). An important feature of our model is that it handles the luminance-dependent processing and spatiallydependent processing independently. This allows us to precompute the expensive spatially-dependent component, making our model extremely efficient. We illustrate the utility of our procedure with global illumination algorithms used for realistic image synthesis. The expense of global illumination computations is many orders of magnitude higher than the expense of direct illumination computations and can greatly benefit by applying our perceptually based technique. Results show our method preserves visual quality while achieving significant computational gains in areas of images with high frequency texture patterns, geometric details, and lighting variations.