RecTangular Multi-Chart Geometry Images

Many mesh parameterization algorithms have focused on minimizing distortion and utilizing texture area, but few have addressed issues related to processing a signal on the mesh surface. We present an algorithm which partitions a mesh into rectangular charts while preserving a one-to-one texel correspondence across chart boundaries. This mapping permits any computation on the mesh surface which is typically carried out on a regular grid, and prevents seams by ensuring resolution continuity along the boundary. These features are also useful for traditional texture applications such as surface painting where continuity is important. Distortion is comparable to other parameterization schemes, and the rectangular charts yield efficient packing into a texture atlas. We apply this parameterization to texture synthesis, fluid simulation, mesh processing and storage, and locating geodesics.

Fast Ray-TRacing of DYNAMIC MESHES USING GEOMETRY IMAGES

Using the GPU to accelerate ray tracing may seem like a natural choice due to the highly parallel nature of the problem. However, determining the most versatile GPU data structure for scene storage and traversal is a challenge. In this paper, we introduce a new method for quick intersection of triangular meshes on the GPU. The method uses a threaded bounding volume hierarchy built from a geometry image, which can be efficiently traversed and constructed entirely on the GPU. This acceleration scheme is highly competitive with other GPU ray tracing methods, while allowing for both dynamic geometry and an efficient level of detail scheme at no extra cost.

PARTICLE BASED FLUID SIMULATION ON THE GPU

Two Algorithms for Fast Reclustering of Dynamic Meshed Surfaces

Numerous mesh algorithms such as parametrization, radiosity, and collision detection require the decomposition of meshes into a series of clusters. In this paper we present two novel approaches for maintaining mesh clusterings on dynamically deforming meshes. The first approach maintains a complete face cluster tree hierarchy using a randomized data structure. The second algorithm maintains a mesh decomposition for a fixed set of clusters. With both algorithms we are able to maintain clusterings on dynamically deforming surfaces of over 100k faces in fractions of a second. (paper). (presentation).

Painting Detail

Surface painting is a technique that allows a user to paint a texture directly onto a surface, usually with a texture atlas: a 1:1 mapping between the surface and its texture image.  Many good automatic texture atlas generation methods exist that evenly distribute texture samples across a surface based on its area and/or curvature, and some are even sensitive to the frequency spectrum of the input texture. However, during the surface painting process, the texture can change non-uniformly and unpredictably and even the best atlases are static and can thus fail to reproduce sections of finely painted detail such as surface illustration.  We present a new texture atlas algorithm that distributes initial texture samples evenly according to surface area and texture frequency, and, more importantly, maintains this distribution as the texture signal changes during the surface painting process.  (paper). (presentation).

Real-Time Subsurface Scattering  

Many semi-transparent objects in the world such as skin, milk, and plants derive some of their visual appearance from how light scatters through them.  Such effects cannot be rendered with the standard local illumination techniques commonly used in real-time computer graphics.  We us a multi-resolution meshed atlas and modern GPU programmability to devise a real-time GPU algorithm that can render semi-transparent objects with diffuse subsurface-scattered illumination under dynamic lighting and viewing conditions.  (paper). (presentation).

The Ray Engine

This research involved the development of a technique for accelerating visibility queries using ray-casting on currently consumer level graphics hardware.  This technique may be used to accelerate a wide variety of applications such as global illumination techniques from Monte Carlo ray-tracing to radiosity.  We developed two models, one using  ATI's Radeon 8000 series, and another more advanced model using NVIDIA's Geforce FX graphics card.  We have received many helpful suggestions from Nvidia on this project.   (paper). (presentation).

General Computing & Graphics Card Programmability

Currently I am working on issues such as efficient hardware graphics programming.  I am in the process of developing a compiler to ease the mapping of high level algorithms to GPU implementations.  The proliferation of extensions to OpenGL has made it difficult to develop programs that work across a variety of graphic card implementations.  As part of this research we've looked at what type of algorithms are naturally adaptable to efficient GPU implementations.  The "Ray Engine" research above is an example of this type of exploration.  

Mip-Mappable Texture Atlas Generation

This research involved examining techniques for parameterizing 3D meshes into the 2D domain.  The 2D domain provides a nice place for both performing computation as well as storing surface properties for the 3D mesh.  This research led us to develop a fully mip-mappable u,v texture space parameterization, that uses 100% of the texels in a square/rectangular texture map resulting in highly effective memory utilization.  This technique proved to be highly effective in minimizing global distortion due to global scaling.  I believe this technique provides an excellent method for parameterization of manifold meshes ( with or without boundary ) for applications such as real-time procedural texturing and 3D painting.   (paper).