Search

Most educators are teaching compter graphics in the same way, and covering the same material, as it was taught 10 years ago. These courses begin with 2D scan-conversion techniques, with 3D topics discussed at the end of the course. This does not take advantage of the great increase in computational and graphics capability of inexpensive computer systems. In this paper, the author proposes a radically different way to approach the introductory course. The new course starts with the basic physics of light-surface interaction, discusses global illumination, and derives the Phong illumination model. The students write ray-tracing programs that incorporate these principles plus shadows and texture mapping. Conventional scan conversion techniques are then discussed later in the course.

Although a great deal of research effort has been devoted to the realistic and efficient rendering of terrain data, the existing research mainly focuses on displaying static terrain. The realistic simulation of dynamic terrain, including vehicle-terrain interactions, is absent in most 3D graphical applications. In this paper, we present a new GPU-based algorithm for dynamic terrain simulation. A key component of this algorithm is a Dynamically-Displaced Height Map (DDHM) which is generated and manipulated on the GPU. In addition, our method achieves real-time performance by using new features of the latest graphics hardware and shader technology. We demonstrate our algorithm by simulating a ground vehicle traveling on soft terrain. Our algorithm is particularly useful for ground based visual simulation applications as well as 3D games.

This poster presents a new technique for visualizing area of interest in a collaborative virtual environment for molecular modeling. The environment allows multiple participants to interact with molecular models while receiving real-time energy feedback from a molecular mechanics simulator. Interaction within the environment occurs primarily via a data glove and is visualized through an HMD (Head-Mounted Display). To interact with the molecule, users define an area of interest by pinching (index to thumb) and then dragging in 3D space. Scientists can then grab within the box to manipulate the molecule. When working with complex molecular models, however, where either the atom count is high, or atoms are clustered together, the area of interest may be occluded by atoms or bonds that fall within the line of site, making it difficult, if not impossible to view the area (Figure 1). While several physically-correct depth-of-focus algorithms have been created [Haeberli and Akeley 1990], we were interested in an efficient approximation specifically designed for this environment.

The introduction of programmable Graphics Processing Units (GPUs) and the addition of procedural shaders to the web3D standard X3D provide us with new techniques to develop real-time Web based visualization. In this paper we discuss the applications of these techniques to bioinformatics and chemistry visualization, specifically the visualization of large biomolecules. By using procedural shaders, we are able to produce higher quality visualizations with minimal performance penalty. We have developed methods to automatically convert from the standard bioinformatics PDB format to CML and then to X3D. The procedural shaders are automatically inserted during the CML to X3D conversion. This provides higher quality images and leads to future possibilities of more flexible and enhanced visualizations.

Graphics processing units (GPUs) have evolved to become powerful, programmable vector processing units. Furthermore, the maximum processing power of current generation GPUs is technically superior to that of current generation CPUs (central processing units), and that power is doubling approximately every nine months, about twice the rate of Moore's law. This research represents the first successful application of GPU vector processing to an existing scientific computing software package, specifically an application for computing the tertiary (3D) geometric structures of protein molecules from x-ray crystallography data. A framework for applying GPU parallel processing to other computational tasks is developed and discussed, and an example of the benefits of taking advantage of the visualization potential of newer GPUs in scientific computing is presented.

The most exciting recent advance in computer graphics has been the development of programmable Graphics Processing Units (GPUs). We discuss different approaches to and some of the issues involved in teaching the use of GPUs.

Each of the three currently available real time shading languages has limitations. The Microsoft HLSL (high level shading language) is Microsoft Windows specific. The NVIDIA Cg shading language is claimed to be platform independent, however our experimental testing results indicated that none of the ATI 9000 level cards could fully support the programs in the Cg toolkit browser [nVidia 2002]. That means we can't say that Cg is really a platform independent shading language. It is still unclear whether OpenGL Shading Language (GLSL) [Rost 2004] can achieve platform independence since it is only supported by a few graphics cards at this point. Until all shading languages achieve platform independence, there is a need to convert shaders in one language to another. Even after this point, the owners of older versions of GPUs will still need a bridge to connect to the newer shader format, especially when they try to publish their shaders on the Internet or load the other version shaders on their graphics card.In this talk, we propose a XML based Universal Converter (UC) for shaders. X3D is defined in XML and the X3D shader group [Carvalho et al. 2003] is working on a shader standard. Thus, we chose to use an XML shader format as the middle layer for the conversion between different shader languages (Figure 1.), so that it will be compatible with any new X3D standard. The UC design, shading languages mapping, difficulties we faced during the implementation and partial implementation results will be discussed.An example of converting a simple vertex shader in GLSL to XML to Cg vertex shader format and vice versa will be presented. The first step is to look at the two corresponding shading languages data-type mapping, structure mapping, qualifier mapping, attribute/semantics mapping, built-in functions mapping, and texture lookup functions mapping. In the presentation we will present mapping tables illustrating these.From looking at the mapping table, it will be apparent that it is not hard to implement the data type conversions. For example, vect2/ vect3/ vect4 and matrix2/ matrix3/ matrix4 in GLSL correspond to the float2/ float3/ float4 and float2x2/ float3x3/ float 4x4 in Cg respectively. We convert ivect/bvect in GLSL to vect type in Cg. That means the middle layer XML will employ the more general data type during parsing.Another issue is mapping attributes in GLSL corresponding to the semantics in Cg. For example, when we parse a structure like: "float4 position: POSITION" in Cg, we need to convert it to a corresponding GLSL built-in attribute "gl_vertex" which is already predefined in GLSL.Some standard library functions in the source language do not have counterparts in the target language. We solve this problem by implementing the "missing" functions using existing functions in the target language. For example,float4 lighting = lit(diffuse, specular, 32); II in Cg would be converted to the following code in GLSL:vec4 lighting; II in OpenGL shading languagelighting.x = 1.0;lighting.y = max(diffuse, 0.0);lighting.z = min(diffuse, speclar);if (lighting. z

A computer graphics course is extremely hardware system dependent, even more than most computer science courses. To produce high quality graphics images requires a high resolution system with extensive color capability and a fast cpu. Fortunately, the computer graphics capabilities of inexpensive systems have continued to increase. As this trend continues we need to consider changing the way we teach our computer graphics courses. In this paper I discuss a major shift in my teaching methods in the past year. Whereas, previously my students developed their own programs to create images, I have switched to the use of the Pixar RenderMan graphics package in the second graphics course and use it at the end of the first graphics course. I will discuss the rational for this change, mine and the students' experiences with it, and future planned modifications of the courses.