Abstract: The benefits of the virtual pixel maps technique in the support of high-quality rendering operations are described. A dedicated frame buffer with a fixed number of bits per pixel is inappropriate for implementing high-quality rendering techniques within the framework of a graphics computer system. The virtual pixel maps feature is an elegant abstraction for solving problems that inherently require a large number of bits per pixel. A system implementation is presented to illustrate the types of hardware rendering algorithms that benefit from the concept of virtual pixel maps. Two specific algorithms, namely, transparency and antialiasing, are used as examples, but many other algorithms could easily be adapted to such an environment. >

Abstract: In medical 3-dimensional display, an input scene is represented by an array of volume elements (abbreviated as “voxels”), and an object in the scene is specified as a “connected” set of voxels. In such applications, surface tracking is an important, and often time-consuming, precursory step. One of the most efficient surface detection algorithms reported in the literature tracks a specified surface of a 3-dimensional object by visiting each boundary face in the surface twice. In this paper, we present a new definition of discrete objects and boundaries that leads to a modified algorithm, which, on the average, visits only one-third of the boundary faces twice (and the rest once). The algorithm has been implemented in our display software package and is found to achieve a run-time reduction of approximately 35%. This timing includes the computation of surface-normal information which is needed for the realistic rendering of surfaces. Without this computation, the saving would be about 55%.

Abstract: This paper introduces the architecture and initial algorithms for Pixel-Planes 5, a heterogeneous multi-computer designed both for high-speed polygon and sphere rendering (1M Phong-shaded triangles...

Abstract: A description is given of a set of algorithms for efficiently rendering an object defined by constructive solid geometry (CSG) directly onto a frame buffer without converting first to a boundary representation. This method requires only that the frame buffer contain sufficient memory to hold two color values, two depth values, and three one-bit flags. The algorithm first converts the CSG tree to a normalized form that is analogous to the sum-of-products form for Boolean switching functions. The following are developed: dynamic interleaving of Boolean tree normalization with bounding-box pruning, allowing efficient rendering for most CSG objects; a method for extending the technique to nonconvex primitives; and implementation of these ideas in an interactive CSG design system on the Pixel-planes 4 solid modeling system. In the design system the designer directly manipulates the CSG structure while continuously viewing the color rendering of the object being designed. >

Abstract: A digital presentation technology that manages anything from text to full-motion video has the potential of expanding the usefulness of personal computers, while rendering them less intimidating.

Abstract: Quaternions, although not well known, provide a fundamental and solid base to describe the orientation of an object or a vector. They are efficient and well suited to solve rotation and orientation problems in computer graphics and animation. This paper describes two new methods for splining quaternions so that they can be used within a keyframe animation system. We also show that quaternions, although up to now solely used for animation purposes, can be used successfully in the field of modelling and rendering and we prove that we can speed up the rendering algorithm by using quaternions.

Abstract: Herein is presented a new procedural method for generating images of trees. Many other algorithms have already been proposed in the last few years focusing on particle systems, fractals, graftals and L-systems or realistic botanical models. Usually the final visual aspect of the tree depends on the development process leading to this form. Our approach differs from all the previous ones. We begin by defining a certain "measure" of the form of a tree or a branching pattern. This is done by introducing the new concept of ramification matrix of a tree. Then we give an algorithm for generating a random tree having as ramification matrix a given arbitrary stochastic triangular matrix. The geometry of the tree is defined from the combinatorial parameters implied in the analysis of the forms of trees. We obtain a method with powerful control of the final form, simple enough to produce quick designs of trees without loosing in the variety and rendering of the images. We also introduce a new rapid drawing of the leaves. The underlying combinatorics constitute a refinment of some work introduced in hydrogeology in the morphological study of river networks. The concept of ramification matrix has been used very recently in physics in the study of fractal ramified patterns.

Abstract: A design is presented for a workstation capable of rendering arbitrary mixtures of analytically defined geometry and sampled scalar fields of three spatial dimensions in realtime or near real-time. The design is based on volumetric compositing, a relatively new approach to visualizing scalar fields by computing 2D projections of a colored semi-transparent volume. Geometric primitives are added by filtering and 3D scan-converting them into the dataset prior to rendering. Speedups totaling four orders of magnitude over published volumetric compositing techniques are obtained through a combination of algorithmic improvements and hardware implementation on Pixelplanes 5, a massively parallel raster display engine incorporating custom logic-enhanced memory chips. A preliminary design is also presented for a two-handed volumeof-interest data exploration tool.

Abstract: We present a technique for modeling global illumination which allows a wide variety of reflectance functions. Scene coherence is exploited in a preprocessing step in which the geometry is analyzed using iterative techniques. Memory is traded for speed, in anticipation of the high memory capacities of workstations of the future. The algorithm operates well over a wide range of time and image quality constraints: realistic results may be produced very quickly while very accurate results require more time and space. The method can be extended for animation and parallelization.

Abstract: The present invention provides unique methods and apparatus for shading curves, polygons and patches, implementing Phong, Gouraud and other shading techniques in the rendering of images on a cathode ray tube or other display device. The present invention also includes a unique method and apparatus for shading patches by rendering a series of adjacent curves such that no pixel gaps exist between each rendered curve.

Abstract: The interactivity provided by rendering at the animation rate is the key to visualizing multivarite time varying volumetric data. We are developing a system for visualizing earth science data using a graphics supercomputer. We propose simple algorithms for real-time texture mapping and volume imaging using a graphics supercomputer. KEYIdORDS." Interactive, texture mapping, volume image, earth science. I N T R O D U C T I O N At the Space Science and Engineering Center we are concerned with the problem of helping earth scientists to visualize their huge data sets. A large weather model output data set may contain one billion points in a five-dimensional array, composed of a 100 by 100 horizontal grid, by 30 vertical levels, by 100 time steps, by 30 different physical variables. Remote sensing instruments such as satellites, radars and lidars produce similarly large data sets. During the last six years we have been developing software tools for managing and visualizing such data sets, as part of the Space Science and Engineering Center's Mancomputer Interactive Data Access System (MclDAS). These tools run on an IBM 4381 and produce animation sequences of multivariate three-dimensional images which are viewed on our large multiframe workstations. However, each image takes 10 to 30 seconds of CPU time to analyze and render, and another 30 seconds to load into the workstation frame store, so turnaround time for producing an animation sequence can be several hours. Figure 1 is a typical image produced by this system. We have applied this system to generate animations from at least twenty different model simulation and remote sensed data sources. Although our earth scientist collaborators are usually pleased with the results, they all express a desire to change the animations with quicker response. This is by far their (and our) primary request. Therefore we have begun developing a highly interactive workstation based on the Stellar GS-1000 graphics supercomputer. This system can produce three-dimensional images from model output data sets in real-time, giving the scientist control over the image generation with immediate feedback. This paper describes earth science data sets, the work we have done so far with the Stellar GS1000, and some thoughts on further development.

Abstract: The authors developed new techniques for three-dimensional display of magnetic resonance (MR) images that preserve soft-tissue definition, are fully automatic, and work with routinely used section thicknesses. MR images are segmented, selectively enhanced, and displayed by means of a volumetric rendering algorithm. These techniques were used to illustrate normal anatomy of the brain, knee, and liver. Three-dimensional rendering of balanced spin-echo images shows the ventricles and extracerebral veins and of T1-weighted images, the sulci and gyri. The large hepatic and portal vessels can be seen with these enhancement techniques. Three-dimensional views of the knee reveal articular surfaces of the tibia and clearly depict menisci and posterior and anterior cruciate ligaments. These techniques make it possible to image multiple soft tissues simultaneously while preserving the detail contained in the original images. Three-dimensional presentation of complex, overlapping anatomic regions is helpful in surgic...

Abstract: A computer graphics technique for rendering shaded areas by using pseudo-random noise localized to image boundaries is disclosed. Calculated pixel values are initially generated using an arithmetic accuracy larger than the word size of an associated bitmap memory. The calculated pixel values thus have a most significant portion, corresponding to the portion which is to be stored in bitmap memory, and a least significant portion, corresponding to the additional accuracy. Pseudo-random noise, preferably generated by a linear feedback shift register, is then added to some or all of the least significant portion, and the carry output from this operation is added to the most significant portion. The most significant portion is then written into bitmap memory. Changes in the value of the integer portion thus occur with increased frequency as the distance to an intensity boundary decreases, since the random noise causes the dithering to occur with increased frequency. The result is a more natural display than would otherwise be available with a given bitmap memory word size.

Abstract: A stereoscopic graphics display system (10) has a stereoscopic window controller (18) that generates multiple windows (72 and 74) within which multiple images (76 and 78) are formed. The stereoscopic window controller directs the graphics displays system to render around each window a border (80 and 82) representing an outline of the window. Each of the borders is rendered with zero binocular disparity to assist an observer to perceive the three-­dimensional qualities of stereoscopic images. Depth cue contradictions between stacked windows are reduced by rendering with zero binocular disparity the images in occluded windows.

Abstract: For most compute environments, adaptive forward differencing is much more efficient when performed using integer arithmetic than when using floating point. Previously low precision integer methods suffered from serious precision problems due to the error accumulation inherent to forward differencing techniques. This paper proposes several different techniques for implementing adaptive forward differencing using integer arithmetic, and provides an error analysis of forward differencing which is useful as a guide for integer AFD implementation. The proposed technique using 32 bit integer values is capable of rendering curves having more than 4K forward steps with an accumulated error of less than one pixel and no overflow problems. A hybrid algorithm employing integer AFD is proposed for rendering antialiased, texture-mapped bicubic surfaces.

Abstract: A method for clipping two and three dimensional graphic primitives for use in a computer graphics workstation. During the first clipping stage, the system removes all graphic primitives which lie outside an arbitrary clipping volume or window that is an enlargement of the desired viewing volume or window. This clipping volume (window) is then projected and mapped onto a virtual viewport which is larger than the real viewport. After this mapping occurs, rendering effects (such as line styles, line width, pattern fill, hatch fill, etc.) are applied to the primitive. A subsequent stage of clipping is then applied to clip the virtual viewport to the real viewport which is the user's visible area. Use of the clipping volume (window) during the first stage of clipping makes it simpler to transform geometric primitives which are partially visible to the viewer, and gives visible primitives a richer set of invariant geometric properties so that their rendering effects may be more easily and correctly carried out.

Abstract: A method and apparatus for tiling a display area defined by lines joining vertices with a surface normal vector associated with each vertex whereby the display area is subdivided into sub-display areas by calculating additional vertices and surface normal vectors by interpolation and rendering a given sub-display area by calculating intensity values at its vertices and tiling its area by linear interpolation of the calculated vertex intensity values.

Abstract: A user, through the use of an input device, provides signals indicative of x,y and z coordinates to a host central processing unit (CPU). The x,y and z coordinates provided by a user correspond to a desired viewpoint for the display of an object stored in an original data image memory. A graphics CPU is coupled to the host CPU for receipt of the x,y and z coordinates as well as to the original data image memory. The graphics CPU is further coupled to a program memory as well as two display buffers, each of which of comprises a bit map of points on the display. The graphics CPU, executing a rendering program stored in the program memory, renders (writes) the data comprising the object for the viewpoint corresponding to the x,y and z coordinates provided by the user. After a predetermined percentage (N) of data comprising the object have been rendered into the non-displayed buffer, the graphics CPU signals a multiplexor which couples the display to the previously non-displayed display buffer, thereby displaying the predetermined percentage (N) of data which have been rendered. In the event that the user does not provide any new x,y and z coordinates representing a different viewpoint, the graphics CPU continues to render the data comprising the object into the displayed buffer, resulting in the appearance to the user that the object is "sparkling" into view. If, however, after the graphics CPU displays the predetermined N percentage of data points, the user provides new x,y,z coordinated corresponding to a different viewpoint, the graphics CPU ceases rendering the object data corresponding to the previous viewpoint, and begins rendering randomized data corresponding to the current viewpoint into the non-displayed display buffer.

Abstract: We present a technique for modeling global illumination which allows a wide variety of reflectance functions. Scene coherence is exploited in a preprocessing step in which the geometry is analyzed ...

Abstract: Today, the shading method of Phong plays an important role in the design of real-time image generation systems. Often, the model has been used in combination with a colour interpolation, suppressing a main property of this model, namely the visually acceptable rendering of highlights. Unfortunately, Phong's algorithm demands a normalisation step, which is expensive to implement in hardware. We will present several shading methods which are reductions of the Phong algorithm. They will be compared both visually and theoretically. The alternatives are judged with respect to their hardware implementation cost. The result is a hierarchy of shading methods that can be used to select the required cost-performance ratio for a given visualisation task.

Abstract: 1 Introduction.- 2 Graphics on general purpose parallel architectures.- 3 A parallel incremental rendering method.- 4 Parallel polygon rendering with precomputed surface patches.- 5 Parallel polygon rendering on a dual-paradigm parallel processor.- 6 Control parallel versus data parallel polygon clipping.- 7 Conclusion.

Abstract: This paper describes a message-based object-oriented tool for exploring mathematically-based interactions which produce complex motions for computer animation. The tool has been implemented as an object in the object-oriented computer animation system The Clockworks. It supports the definition of complex interactions between geometric objects through the specification of messages to the interacting objects. Our approach is general, flexible and powerful. The tool itself is not hardcoded to a particular application. It simply sends the messages specifies by the user. Messages are specified as strings which may be stored, modified and interpreted. Since the tool is part of the Clockworks it may utilize many of the powerful features of the system, including data structuring, mathematical, geometric modeling, and rendering objects. The tool has been used to explore a general spring and mass model, and the response of objects in a vector field.(AGD)

Abstract: Medical workstations of the future will support the real-time display and interactive manipulation of 3-D objects derived from CT, MRI and other imaging modalities. As part of such an integrated system for visualizing 3-D volumes, we have developed an highly interactive, flexible, and portable program for the 2D display of image slices and contours outlining anatomical objects. The editing of these contours as well as their automatic creation by thresholding and edge-tracking is supported. The contours may later be used to generate 3D surfaces for shaded-graphics rendering, or to mask out regions of interest in the image for volume rendering. This "image executive" program, or Imex, is designed to run in a windowing environment (i.e., the X Window System). The user-interface model, which may be described as a "Macintosh 1 for images", associates one movable and resizable window with each displayed view of a 2D image slice or of an indexed array of slices. Any number of views into one slice or into a subset of an array of slices may be present. Natural interaction is achieved by providing immediate response and by using the mouse to effect navigation and viewing functions, for example grabbing and dragging a slice onto another to copy its field-of-view or other attributes.

Abstract: We describe a method for matching computed tomography (CT) and/or magnetic resonance (MR) medical image data with digital subtraction angiograms (DSA). The tomographic data set is processed and rendered to yield a volumetric image which is displayed stereoscopically from the same viewpoint at which the DSA images were acquired. The vessels are merged on top of the 3D rendered volumes and the composite images are viewed stereoscopically on a PC-based workstation. The workstation has facilities allowing the user to manipulate a. cursor or vector in stereo within a defined 3D coordinate system and to make accurate 3D measurements within the imaged volume. While surgical planning can be conducted from analysis of the separate tomographic images and individual bi-plane angiograms, such an approach is tedious since as many as 400 separate images from separate modalities may be acquired for a single patient. Furthermore, it is sometimes difficult to fully appreciate the anatomy of the surrounding structures from individual CT and MR slices of the 3D volume. All image processing and rendering is performed on a PIXAR Image Computer, a SIMD (single instruction, multiple data) channel array processor with a 48 - bit deep image buffer. PIXAR's Chap Volumes rendering library is employed in the rendering process. The stereoscopic workstation consists of a high performance color monitor which displays the left and right eye views of the data sequentially at 120 Hz via a liquid crystal polarizing shutter. The operator views the image via left and right eye circularly polarized glasses.

Abstract: Traffic safety is highly correlated to the amount of visual information that can be obtained from the road and its immediate environment. It is therefore a logical consequence to base any quality judgment for lighting systems on visibility criteria. This development can be observed in indoor lighting as well, where the visibility-related CRF (contrast rendering factor) was introduced as a quality criterion. Visibility as a characteristic of roadway lighting has recently been discussed in Canada. When applying visibility as a criterion, there needs to be a metric to measure it and a method for calculation to predict the visibility level to be achieved in a certain lighting installation.

Abstract: This paper proposes an efficient method for rendering highlights on rounded edges, which is important for photorealism and comprehensibility. The rounded edges are shaded as thin cylinders separately from the planar surfaces. To ensure coherence with the planar surfaces, an edge shading equation is proposed, which derives an appropriate edge shading model from any conventional model. The final image is obtained by drawing edges like wire-frames onto the planar surface image. Using this method, aliasing-free edge highlights can be generated from simple edge data with little increase in computation cost.

Abstract: Tools specifically oriented toward flow analysis problems are described. The approach is oriented toward performing graphics calculations on the supercomputer and using high-end workstations as rendering engines to display flow fields generated by aircraft flight. A suite of software provides both display and interactive analysis capability. Also described is animation hardware featuring Winchester disk technology, which allows the storage of more than one gigabyte of data and supports editing, special effects, and video output. >

Abstract: Apparatus is provided for rendering uniform a static magnetic field which includes a plurality of radial and axial coils connected to a controlled power supply. This apparatus permits control of at least the first six orders of gradient of the static magnetic field.

Abstract: The architecture and the hardware realization of the 3D Viewing and Rendering Processor is presented. This processor is a component of the Cube architecture, developed primarily for volume visualization. The processor generates 2D shaded orthographic, parallel, and perspective projections of the volumetric image of n3 voxels in O(n2log n) time. This performance is attributed to a unique skewed memory organization, a special ray projection bus, an extended viewing architecture, and a new congradient shading technique. A reduced-resolution prototype has been realized in hardware using printed circuit board technology and has been running in true real time. Currently, a VLSI version of the prototype is being tested.