🚀 go-pugleaf

The following are abstracts of recent technical reports from the
University of North Carolina.  To receive hard copies, contact
Research Access, Inc. at 1-800-685-6510 or 412-682-6510.  To
receive the reports electronically, send email to "netlib@cs.unc.edu",
and in either the subject line or message body, send the request,
"get <report> from techreports".  These files can be read with
Microsoft Word or WriteNow. 

If you have any questions/problems, please send mail to softlab@cs.unc.edu
___________________________________________________________________________


Design and Construction of a 30-Degree See-Through Head-Mounted Display
TR92-030
Douglas E. Holmgren
**(This report is available electronically by emailing netlib@cs.unc.edu
   and sending the request "get 92-030.rtf from techreports")
This report describes the design and construction of a 30-degree see-through
head-mounted display (STHMD).  The design goals for this STHMD were that it
employ calibrated means of adjusting the virtual image distance (focus) and
inter-pupillary distance (IPD) and provide for adjustment of exit pupil
location with respect to individual wearers.  The optics were designed by
Jannick Rolland to incorporate off-the-shelf optical components.


Scanned Laser Displays for Head Mounted Displays
TR92-029
Douglas E. Holmgren, Warren Robinett
**(This report is available electronically by emailing netlib@cs.unc.edu
   and sending the request "get 92-029.rtf from techreports")
Technologies applicable toward a display system in which a laser is 
raster scanned on the viewer's retina are reviewed.  The properties 
of laser beam propagation and the inherent resolution of a laser 
scanning system are discussed.  Scanning techniques employing 
rotating mirrors, galvonometer scanners, acousto-optic deflectors, 
and piezo-electric deflectors are described.  Resolution, speed, 
deflection range, and physical size are strongly coupled properties of 
these technologies.  For head-mounted display applications, a 
monochromatic system employing a laser diode source with acousto-
optic and galvonometer scanners is deemed most practical.  A 
resolution of 1000 x 1000 pixels at 60 frames per second should be 
possible with such a sytem.  A radiometric analysis indicates that eye 
safety would not be a problem in a retina-scanning system.

Six Generations of Building Walkthrough: Final Technical Report to the
National Science Foundation
TR92-026
Frederick P. Brooks, Jr., John Airey, John Alspaugh, Andrew Bell, 
Randolph Brown, Curtis Hill, Uwe Nimscheck, Penny Rheingans, John Rohlf,
Dana Smith, Douglass Turner, Amitabh Varshney, Yulan Wang, Hans Weber,
Xialin Yuan
**(This report is available electronically by emailing netlib@cs.unc.edu
   and sending the request "get 92-026.rtf from techreports")
NO ABSTRACT FOR THIS REPORT.

Synthetic Experience:  A Taxonomy, Survey of Earlier Thought, and
Speculations on the Future
TR92-022
Warren Robinett
**(This report is available electronically by emailing netlib@cs.unc.edu
   and sending the request "get 92-022.rtf from techreports")
A taxonomy is proposed to classify all varieties of technologically-
mediated experience.  This includes virtual reality and teleoperation, 
as well as earlier devices such as the microscope and telephone.  The 
model of mediated interaction assumes a sensor-display link from 
the world to the human, and an action-actuator link going back from 
the human to the world, with the mediating technology transforming 
the transmitted experience in some way.  The taxonomy is used to 
classify a number of sample systems.

Earlier visions of technologically-based experience are compared 
with the ideas presented in this paper. Then the long-term prospects 
of this field are speculated upon, ignoring constraints of cost, effort, 
or time to develop.  Finally, the ultimate limits of synthetic 
experience are discussed, derived from properties of the physical 
universe and the human neural apparatus.


A Survey of Data Models for Hypermedia
TR92-021
Joan Boone
This article surveys data models developed in support of
hypermedia environments.  Each model is briefly summarized
and identified according to the primary motivating factor:
application requirements, refinement and formalization, and
most recently, standardization.  A more detailed description
of the following data model characteristics is provided:
low-level abstractions, high-level abstractions, operations,
and relationship to system architecture.  Lastly, the models
are reviewed with respect to how  they address some
hypermedia issues such as versioning, collaboration,
interoperability and interchange, and security.


A Distributed Implementation of an N-body Virtual World Simulation
TR92-020
Mark Parris, Carl Mueller, Jan Prins, Adam Duggan, 
Quan Zhou, Erik Erikson
**(This report is available electronically by emailing netlib@cs.unc.edu
   and sending the request "get 92-020.rtf from techreports")
Meeting the challenge of real-time visualization of a complex 
physical simulation can require several specialized computers.  The 
virtual-world simulation described in this paper combines a highly-
parallel computer for physical simulation and a custom graphics 
engine for display communicating via TCP/IP over an ethernet.  
Application-level protocols to provide the high-throughput, low-
latency communication required to support an interactive virtual 
world are investigated.  Experiments demonstrate that a limited 
write-ahead protocol gives the best tradeoff between latency and 
throughput for this application.


A Short Guide to AutoCAD Drawing Primitives for 3D Computer Graphics Models 
and the Walkthrough AutoCAD-to-Polygon Conversion Program 
TR92-019
John C. Alspaugh
**(This report is available electronically by emailing netlib@cs.unc.edu
   and sending the request "get 92-019.rtf from techreports")
Intended for those who want to build models for 3D computer 
graphics as painlessly as possible, this document introduces the 
reader to those AutoCAD drawing primitives and AutoCAD structures 
that can be converted to polygons with the Walkthrough AutoCAD-
to-polygon conversion program.  It also discusses those primitives 
that are not converted and why they are not converted.  In addition, 
this document introduces the reader to more advanced AutoCAD 
modelling techniques for efficient model-building and model-
management.  The final section tells how to run the AutoCAD-to-
polygon conversion program:  file locations, input file requirements 
and forms, and the format of the output file.

Implementation of Randomized Sorting on Large Parallel Machines
TR92-018
William L. Hightower, Jan F. Prins, John H. Reif
**(This report is available electronically by emailing netlib@cs.unc.edu
   and sending the request "get 92-018.rtf from techreports")
Flashsort [RV83,86] and Samplesort [HC83] are related parallel 
sorting algorithms proposed in the literature.  Both utilize a 
sophisticated randomized sampling technique to form a splitter set, 
but Samplesort distributes the splitter set to each processor while 
Flashsort uses splitter-directed routing. 

In this paper we present B-Flashsort, a new batched-routing variant 
of Flashsort designed to sort N>P values using P processors connected 
in a d-dimensional mesh and using constant space in addition to the 
input and output. The key advantage of the Flashsort approach over 
Samplesort is a decrease in memory requirements, by avoiding the 
broadcast of the splitter set to all processors.  The practical 
advantage of B-Flashsort over Flashsort is that it replaces pipelined 
splitter-directed routing with a set of synchronous local 
communications and bounds recursion, while still being 
demonstrably efficient.

The performance of B-Flashsort and Samplesort is compared using a 
parameterized analytic model in the style of [BLM+91] to show that 
on a d-dimensional toroidal mesh B-Flashsort improves on 
Samplesort when (N/P) < P/[(c1)(log P)+(c2)(d)(P^(1/d))+(c3)], for machine-
dependent parameters c1, c2, and c3.  Empirical confirmation of the 
analytical model is obtained through implementations on a MasPar 
MP-1 of Samplesort and two B-Flashsort variants.


Spectra-Limited Diffusion in the Characterization of
Textured Patches in Digital Images
TR92-017
Ross Whitaker
NO ABSTRACT FOR THIS REPORT.
REPORT NOT YET AVAILABLE.


Techniques for Interactive Manipulation of Graphical Protein Models 
(Under the direction of Frederick P. Brooks, Jr. ) 
TR92-016 
Mark Christopher Surles 
**(This report is available electronically by emailing netlib@cs.unc.edu and 
   sending the request "get 92-016.sit from techreports")
This thesis describes a graphics modeling system, called Sculpt, that 
maintains physically-valid protein properties while a user 
interactively moves atoms in a protein model.  Sculpt models strong 
properties such as bond lengths and angles with rigid constraints and 
models weak properties such as near-neighbor interactions with 
potential energies.  Sculpt continually satisfies the constraints and 
maintains a local energy minimum throughout user interaction.  On a 
Silicon Graphics 240-GTX, Sculpt maintains 1.5 updates per second on 
a molecular model with 355 atoms (1065 variables, 1027 constraints, 
and 3450 potential energies).  Performance decreases linearly with 
increased molecule size.  Three techniques yield interactive 
performance:  a constrained minimization algorithm with linear 
complexity in problem size, coarse-grain parallelism, and variable 
reduction that replaces model segments with rigid bodies.

The thesis presents a Lagrange multiplier method that finds a 
constrained minimum and achieves linear computational complexity 
for articulated figures whose spine contains many more joints than 
any attached limb (e.g. reptiles, mammals, and proteins).  The 
method computes the Jacobian matrix of the constraint functions, 
multiplies it by its transpose, and solves the resulting system of 
equations.  A sort of the Jacobian at program initialization yields a 
constant, band-diagonal pattern of nonzeros.  Multiplication and 
solution of band-diagonal matrices require computation that 
increases linearly with problem size.  One or two iterations of this 
algorithm typically find a constrained minimum in this application.

The number of functions and variables can be reduced by the use of 
rigid bodies.  A user can specify that a rigid object with few variables 
replace large segments of a model that should not change.  For 
example, a user can twist a backbone into a helix and then freeze the 
helix by replacing its atoms and bonds with a cylinder of rigid shape 
but movable position and orientation.

Two improvements over existing interactive protein modeling 
systems have been observed in modeling sessions with Sculpt.  First, 
time-consuming model correction is avoided by maintaining a 
physically-valid model throughout a modeling session.  Second, 
additional cues about model properties can arise when a chemist 
interactively guides a folding simulation rather than viewing a cine 
loop from a pre-computed simulation.


Eighteenth Annual Progress Report:  Interactive Graphics for
Molecular Studies
TR92-015
Frederick P. Brooks, Jr., William V. Wright
**(This report is available electronically by emailing netlib@cs.unc.edu
   and sending the request "get 92-015.rtf from techreports")
NO ABSTRACT FOR THIS REPORT


Adaptive Mesh Generation for Radiosity Methods
TR92-014
Uwe M Nimscheck
**(This report is available electronically by emailing netlib@cs.unc.edu
   and sending the request "get 92-014.rtf from techreports")
The radiosity method is one of the main approaches to 
the generation of realistic images.  The generation of 
accurate radiosity solutions imposes numerous 
constraints on the input data.  Generally, a radiosity 
pipeline can be subdivided into modeling, preprocessing, 
mesh generation, radiosity solution and display.  This 
thesis investigates the modeling requirements for valid 
radiosity solutions, reviews existing mesh generation 
algorithms for finite element methods, and introduces a 
new mesh generation algorithm for interactive radiosity 
approaches.  The algorithm consists of a spatial 
decomposition by a modified quadtree and a subdomain 
meshing by removal of individual patches.  It runs in 
O(n) time, with n being the number of generated patches.  
The use of a tree structure allows one to increase and 
decrease the density of the generated mesh locally and 
makes the mesh generator capable of adapting the mesh 
to the current lighting situation.  This makes the 
proposed algorithm suitable for adaptive refinement and 
interactive radiosity solutions.


The Constitution of Intellectual Work By Programming Teams
TR92-013
Dorothy Holland, James R. Reeves and Anne Larme
This paper is drawn from a large study of intellectual
collaboration.  In a component of the research we focused on
computer scientists and the ways in which they work together
to produce complex programs.  Relying upon the methods of
participant-observation and analytic-induction we followed
four teams of programmers from the inception of a project to
its completion.  Here we present our analyses of the groups'
intellectual work as situated in cultural systems, social
relations, and institutional matrices.  In particular we
describe the teams' changing perspectives on the purpose of
their work.  These perspectives which we call "activity
frames" affected the type of work that the teams did and the
manner in which it was carried out.  Our findings have
implications for the building of computer tools, and for the
theorizing and researching of collaborative work.


Architecture of the Artifact-Based Collaboration
System Matrix
TR92-012
Kevin Jeffay, Jin-Kun Lin,  John Menges, F. Donelson
Smith, and John B. Smith
The UNC Collaboratory project is concerned with both the
process of collaboration and with computer systems to support
that process.  Here, we describe a component of the Artifact-
Based Collaboration (ABC) system, called the Matrix, that
provides an infrastructure in which existing single-user
applications can be incorporated with few, if any, changes
and used collaboratively.  We take the position that what is
needed is not new  tools but better infrastructure for using
familiar single-user tools collectively.  The paper discusses
the Matrix architecture, a Virtual Screen component, and
generic functions  that provide conferencing, hyperlinking,
and recording of usersU actions for all applications.

Collective Intelligence in Computer-Based
Collaboration:  An Introduction
TR92-011
John B.  Smith
A Concept of Collective Intelligence  in computer-based
collaboration is defined.  It views Collective Intelligence
as the goal groups try to achieve when they seek to meld
their thinking into a coherent whole--as if the result were
the work of a single good mind. Components are identified
within computer-supported collaborative groups that are
recognizable as extrapolations of analogous components in
models of individual cognition.  These include several types
of memories, processors, and executive functions.  Several
dimensions are also discussed in which to locate studies,
systems, and issues.  The paper suggests a broad agenda for
collaboration research.


A Distributed Graph Storage System for Artifacts in
Group Collaborations
TR92-010
Douglas  Shackelford, John B. Smith and F. Donelson Smith
Our group is building an artifact-based collaboration support
system and studying the collaborative process.  This paper
discusses the data storage component of the system.  It
supports a graph-based data model, conservatively extended to
meet hypermedia requirements.  It is implemented in a
distributed architecture so data may be stored in multiple
locations and moved among locations .  Issues addressed in
both the paper and the system include: scale, logical and
physical partitioning, protection, concurrency, and support
for an extensible set of user applications.  The discussion
emphasizes issues and system aspects concerned with
collaboration and support of multiple concurrent users.


Re-Tiling of Polygonal Surfaces
TR92-006
Greg Turk
**(This report is available electronically by emailing netlib@cs.unc.edu
   and sending the request "get 92-006.rtf from techreports")
This paper presents an automatic method of creating 
surface models at several levels of detail from an original 
polygonal description of a given object.  Representing 
models at various levels of detail is important for 
achieving high frame rates in interactive graphics 
applications and also for speeding-up the off-line 
rendering of complex scenes.  Unfortunately, generating 
these levels of detail is a time-consuming task usually 
left to a human modeller.  This paper shows how a new 
set of vertices can be distributed over the surface of a 
model and connected to one another to create a re-tiling 
of a surface that is faithful to both the geometry and the 
topology of the original surface.  The main contributions 
of this paper are:  1) a robust method of connecting 
together new vertices over a surface,  2) a way of using 
an estimate of surface curvature to distribute more new 
vertices at regions of detail.  The key notion in the re-
tiling procedure is the creation of an intermediate model 
called the mutual tessellation of a surface that contains 
both the vertices from the original model and the new 
points that are to become vertices in the re-tiled surface.  
The new model is then created by removing each original 
vertex and locally re-triangulating the surface in a way 
that matches the local connectedness of the initial 
surface.  This technique for surface re-tessellation has 
been successfully applied to iso-surface models derived 
from volume data, Connolly surface molecular models and 
a tessellation of a minimal surface of interest to 
mathematicians.  Polygonal models at multiple levels of 
detail can be put to immediate use in interactive graphics 
applications such as radiation treatment planning.


Merging Virtual Reality with the Real World:
Seeing Ultrasound Imagery within the Patient
TR92-005
Mike Bajura, Henry Fuchs, Ryutarou Ohbuchi
**(This report is available electronically by emailing netlib@cs.unc.edu
   and sending the request "get 92-005.rtf from techreports")
We describe initial results showing "live" ultrasound 
echography data visualized within a pregnant human 
subject.  The visualization is achieved using a small video 
camera mounted in front of a conventional head-mounted 
display worn by an observer.  The camera's video images 
are composited with the computer generated ones that 
contain one or multiple 2D ultra-sound images properly 
transformed to the observer's current viewing position. 
As the observer walks around the subject, the ultrasound 
images appear stationary in 3-space within the subject.  
This kind of enhancement of the observer's vision may 
have many other applications, both in medicine - seeing 
arteries to avoid that lie near a pending surgical incision - 
and in other areas - seeing inside burning buildings with 
hand-held radar or walking inside house designs on site.


Viper:  A Quasi-Real-Time Virtual-Worlds Application
TR92-004
Richard L. Holloway
Virtual-worlds applications have a great need for 
interactive performance, since the user's view of the real 
world may be blocked out and the only "world" the user 
can see is the virtual world created by the application. 
For this reason, the system's response is critical-  if the 
user turns his head and the world takes seconds to catch 
up with him, the result will be disorienting and confusing, 
rather than enlightening.  The viper application is an 
attempt at giving real-time or near-real-time 
performance in a virtual-worlds application;  it allows the 
user to trade off image quality for interactive 
performance.


Designing Theory-Based Systems:  A Case Study
TR92-003
John B. Smith and Marcy Lansman
In this paper, we discuss principles for designing and
testing computer systems intended to support usersU thinking
as they perform open-ended or ill-defined tasks.  We argue
that such systems inherently and inevitably implement a model
of usersU cognitive behaviors.  Making that model explicit
can provide system developers with guidance in making design
decisions.  However, both model and system must be tested and
refined.  We discuss these principles in relation to a case
study in which our group developed a hypertext-based writing
environment and then tested that system in a series of
experimental studies of writers' strategies.


Comprehensive C++ I/O Libraries Supporting Image 
Processing in a University Research Environment
TR92-002
Alfred Graham Gash, Fred J.R. Appelman, 
Karel J. Zuiderveld
**(This report is available electronically by emailing netlib@cs.unc.edu
   and sending the request "get 92-002.tex from techreports")
An on-going effort to develop a set of class-based 
libraries in C++ for image I/O and image processing is 
described.  These new libraries will provide the following 
features to application programs based on them:

	Derivable image classes
	Machine-architecture independence of data storage
	Fast, user-customizable image I/O
	Unlimited number of descriptive data items
	Arbitrary number of image dimensions
	User-definable data types
	Optional automatic type conversion and 
	compression
	Support for accessing images using reference files
	  and pipes

This highly-flexible and powerful capability is designed 
to meet the needs of workstation-based image processing 
and computer vision research in a multi-departmental 
university environment.


The Study of Context Effects in Medical Image Contrast 
Enhancement Assessment via a Binocular Ganzfeld
TR92-001
Derek T. Puff
The context in which local structures in a medical image 
are viewed has an impact on a human observerUs ability to 
localize, identify, and characterize those structures and as 
such is an important visual feature for incorporation in 
models and methods in medical image enhancement and 
display.  Medical image contrast enhancement methods 
specify a transformation by which acquired image values 
are mapped to the display in a way that increases the 
contrast between relevant anatomical structures and the 
background or adjacent structures.  These contrast 
enhancement techniques may eventually be evaluated 
with computer methods that predict the performance for 
an array of modeled visual tasks.  The model utilized to 
define the visual mechanisms required in these tasks 
must incorporate a knowledge about particular influences 
of context on localized detection and discrimination.  In 
pursuit of this understanding, several context 
experiments were conducted with a binocular ganzfeld.  
The physical characteristics of the ganzfeld, which allows 
presentation of greyscale images without the influence of 
extraneous surrounding contextual structures of borders 
present under normal display conditions, and the results 
of those studies, are presented.



===========================================================================
Co-moderator:  Richard Golding, Computer & Information Sciences, UC Santa Cruz
	    compdoc-techreports-request@ftp.cse.ucsc.edu