|
Distributed Applications and Systems LabDirector - Prof. Ray Ford , Department of Computer Science |
Project Descriptions
For more details on any of the projects described below, consult the references listed under "Publications" or send e-mail to ford@cs.umt.edu.
1. Ecosystem Information System (EIS)
PIs: Ray Ford/DASL, Steve Running/NTSG, Rama Nemani/NTSG,
Roly Redmond/WSAL, Mike Sweet/Forestry
NSF - Database Activities in BBS Program, June 1993 - December 1996
McIntire-Stennis Cooperative Research Program, January 1994 - December 1996
The EIS Project involves the design and implementation of an Internet-accessible information system which can serve as a repository of datasets and models accessible to the world-wide community of ecosystem researchers. EIS will provide free and easy access to a wealth of existing datasets and models, and will allow researchers to also create local repositories of their own datasets and models which can then be "registered" in a global "web" to become part of the globally accessible "web" of information. The EIS web of local and global information is patterned after that of World-Wide-Web, which you are currently using. The EIS also includes a World-Wide-Web interface, which allows Internet users to browse EIS sites connected to the WWW server host. EIS uses object-oriented technology in its characterization of the ecosystem database -- the database is organized as a class hierarchy with associated instances (datasets) and methods (executable data transformations). The EIS design and implementation is also based on object-oriented technology, i.e., beginning with an object-oriented design and continuing through a C++ client/server implementation.
2. Petroleum Reservoir Visualization
PIs: Chuck Wideman/Geophysics-Montana Tech, Ray Ford/DASL, ... (many others) DOE - EPSCOR Program, October 1993 - September 1999.
The Petroleum Reservoir Characterization project is a multi-investigator, multi-campus project involving geologists, geophysicists, and computer scientists from Montana Tech, the University of Montana, and Montana State University. The project seeks to develop novel techniques for the characterization of petroleum reservoirs, based on novel analysis of 3-D seismic data, production/treatment modelling, and visualization of available data. DASL is responsible for the visualization component, and is developing visualization tools in two forms: calibration support tools as used by the modeller in the interpretation and tuning of models early in the research process, and presentation tools used late in the research process to present research results to a wider audience.
3. IBM Data Explorer Interface/Repository and OpenDX Consortium .
PI: Ray Ford/DASL IBM - Visualization Group, June 1993 - June 1999.
For the past several years DASL has been working with the IBM Visualization Group to expand the range of application-specific tools available for use with IBM's Data Explorer ( DX ) visualization software. DASL's focus continues to be on applications related to environmental modeling and natural resource management. DASL supports a repository of tools and demonstations that illustrate how to easily import datasets in standard data formats (digital elevation maps, digital line graphs, ERDAS/Lan files, etc.). Current work focuses on integrating widely used modelling systems (e.g., USGS ground-water modeling system Modflow) into DX, and building data interfaces with commercial geographic information systems (e.g., Arc/Info).
During the summer of 1999 IBM went "open source" with the DX code, prompting users at various sites, including DASL, to form the OpenDX Consortium to assure continued coordinated development and support for this code. DASL's David Thompson is one of the OpenDX co-founders, and continues to play a leading role in charting DX/OpenDX development.
4. Image Analysis and Aggregation Algorithms
PI: Ray Ford/DASL
Collaborating Scientists: Roly Redmond/WSAL, Marc Hendrix/Geology
DOE and a collection of natural resource management agencies,
January 1993 - ongoing
For several years DASL staff has been working with application scientists in other disciplines on the development of specialized algorithms for the classification and analysis of very large scale imagery. Some examples are
-
Aggregation in the Classification of Landcover, with the
University
of Montana Wildlife Spatial Analysis Lab, Director: Dr. Roly Redmond.
The major focus of WSAL work has been the classification of land cover, habitat, and other ecosystem parameters for large regions and at a variety of scales. The basic inputs to the classification process are satellite imagery, which provides information at a very fine resolution (30m x 30m). DASL staff have been working with WSAL application experts to develop efficient, rule-based algorithms that maintain the integrity of the data as the resolution is dropped to the coarser levels necessary to describe very large regions. Algorithms have been developed that will process large areas (roughly 8000 x 8000 30m square pixels) in about an hour on a modestly configured Unix workstation. Ground truthing of the resulting imagery is currently underway. Results of ground truthing will be used to fine-tune current algorithms, and to develop the next round of algorithms that adjust the data to even coarser scale. - Petrographic Image Analysis,
with Dr. Marc Hendrix, UM Department of Geology and Dr. Chuck Wideman,
Montana Tech Department of Geophysics
The focus of this work is the analysis of very large scale imagery from thin sections of oil well cores. The purpose is automated estimation of key parameters in oil well production, namely porosity and permeability. - "Giga" and "Tera" scale image analysis,
with Steve Barsness
The focus of this work is on the development of algorithms that scale to the analysis of images on the order of 10K x 10K, 100K x 100K, or larger, following a path of development initiated by Steve Barsness while he was a graduate student/DASL researcher. Steve has formed a company company to make his initial version of this software available, but much research remains to be done on the development of theory and practical algorithms in this area.
5. Visualization and Multimedia Presentations for Natural Resource Applications
with: Steve Running/NTSG, Roly Redmond/WSAL, DOE Petroleum Reservoir
Characterization Project, and others
NASA, US Forest Service, USGS, DOE, and others, June 1994 - ongoing
DASL visualization experts continue to work with scientists in various natural resource oriented research labs, and with students and staff in the University of Montana's media arts program, to develop various types of multimedia presentation materials used to describe the science to other scientists, funding agencies, and the public. Examples include static imagery and "Web movies" on some of the Web sites noted above, as well as instructional CDs and video tapes. Examples of this work is available on request.
6. MMS/DX Interface Development
PI: Ray Ford and Dick Thompson/DASL IBM - GIS Group, March 1994 - December 1994.
DASL is currently working with a consortium of government, university and commercial partners known as TerraLab in the development of a version of the USGS's Modular Modeling System ( MMS ) that will contain embedded links to a GIS (GRASS), a visualization package (DX), and a collection of modelling and analysis modules. DASL's responsibility is building the MMS/DX interface.