|
|
|
Space-Optimized Texture Maps:
We propose a new texture optimization algorithm based on the reduction
of the physical space allotted to the texture image. The resulting
image can be resampled at lower rate while preserving its original
details. We present several examples showing that our method
significantly decreases texture memory usage without noticeable loss
in visual quality.
|
|
|
Deep View Visualization System:
A recent trend in the graphics and visualization communities takes
advantage of the significant processing power available in today's
commodity processors and graphics adapters. By combining the power of
many off-the-shelf products, very large data sets can be managed using
parallel processing. Along these lines, our Deep View visualization
system combines the processing, graphics, and networking power of a
cluster of workstations to provide an inexpensive, parallel rendering,
and video streaming system.
|
|
|
Modeling
with Subdivision Surfaces: We explore the applicability of
subdivision techniques to interactive surface editing for shape
design. We have focused on the implementation of an interactive
cut-and-paste tool for surfaces, a method for creating sharp features
and trim regions on multiresolution subdivision surfaces, and a method
of deforming a geometric shape so that is satisfies various
constraints. |
|
|
3D Scanning: Three dimensional scanning has recently become
a very active area in computer graphics. The requirements for computer
graphics are different from those of traditional scanning
applications. We are developing scanning systems for producing
virtual objects that can be rendered with high visual quality (e.g.,
see Michelangelo's Pietà project below).
|
|
|
Simplification of Surface Annotations:
Geometric models are often annotated to provide additional information
during visualization. Texture maps have often been used to convey
this information at relatively low cost, but they suffer from blurring
and pixelization at high magnification. We present a technique for
simplifying surface annotations based on directed, asymmetric
tolerance. By maintaining the annotations as geometry, as opposed to
textures, we are able to simplify them while still maintaining the
overall appearance of the model.
|
|
|
Multiresolution Subdivision Surfaces:
We investigate data structures and algorithms for efficient
representation and transmission of multiresolution subdivision
surfaces. This work builds on previous results for the representation
of multiresolution terrains.
|
|
|
Michelangelo's Pietà:
We have been engaged in a project to create a detailed
three-dimensional model of Michelangelo's Florentine Pietà. This
project was defined and driven by the research interests of Dr. Jack
Wasserman, professor emeritus of art history at Temple University. We
used a 3D scanner to measure shape and appearance of each part of the
statue, and developed new algorithms to merge all the information into
a single accurate digital model.
|
|
|
Pervasive 3D Graphics: We are evolving towards a future in
which access to information will be pervasive. New devices to browse
and interact with this information are appearing. We are experimenting
with rendering and interaction techniques suitable for this new class
of devices.
|
|
|
3D Java Viewer: We have developed a pure-Java player that
combines our 3D rendering engine with MPEG-4 3D mesh compression and
coding. The result is a lightweight, software-only client with good
performance both in terms of transmission time and frame rate. The
viewer is part of IBM's award winning HotMedia multimedia toolkit for
the web.
|
|
|
Prioritized-Layered Projection (PLP):
We present a technique for optimizing the rendering of high-depth
complexity scenes by rendering an estimation of the visible set for
each frame. The novelty in our work lies in the fact that we do not
explicitly compute visible sets. Instead, our work is based on
computing on demand a priority order for the polygons that maximizes
the likelihood of rendering visible polygons before occluded ones for
any given scene.
|
|
|
Adaptive Rendering and Transmission Environment (ARTE):
Emerging networking infrastructures include an increasing variety of
clients and servers interconnected by communication fabrics of various
types and capabilities. Efficient delivery of 3D graphics over
networks must adapt to both network resources and capabilities of
various client platforms. To address these problems we have designed
and implemented ARTE.
|
|
|
OpenDX: A visualization framework that gives users the
ability to apply advanced visualization and analysis techniques to
their data. These techniques can be applied to help users gain new
insights into data from applications in a wide variety of fields. DX
provides a full set of tools for manipulating, transforming,
processing, realizing, rendering and animating data and allow for
visualization and analysis methods based on points, lines, areas,
volumes, images or geometric primitives in any combination.
|
|
|
Pro/3000 Digital Imaging System: To capture archival
quality high resolution images from transparencies, reflective media,
or three dimensional objects, we have designed and built the Pro/3000
Digital Imaging System. The Pro/3000 consists of a highly versatile
three-pass color digitizing scanner, supports formats from 35mm to 24
by 30 inches, and is easy to use with its camera-like features such as
the "through-the-lens viewfinder" which provides accurate focusing and
framing.
|
|
|
Other projects in