CE691M - Geomatics Engineering Seminar, 2006 Fall

Time: 11:30 – 12:20, Thursdays; Room: CIVL 2118

Nov. 30, Tom Lobonc, Ph.D, Leica Geosystems

Challenges and Solutions in the Geospatial Industry

In academic pursuits, research activities are usually centered around advancing algorithmic state of the art, and student exposure to a broad set of data and problems can be limited. One key to developing the most effective geospatial tools is to understand the challenges in a functional production and exploitation environment. The presenter will share his experiences in the geospatial industry, discuss problems and solutions, and review needs and trends.

Dr. Lobonc is currently serving in the role of Director of Photogrammetry and Defense Product Lines for Leica Geosystems Geospatial Imaging in Norcross, Georgia. In this capacity he is responsible for identifying market needs and trends and setting the direction for Leica products. He also provides direction to the Leica defense engineering team and contributes photogrammetric expertise on various projects. Prior to joining Leica, Dr. Lobonc spent nine years with Science Applications International Corporation as a Senior Photogrammetrist and project manager. In addition to performing sensor modeling and algorithm design, he was a key individual in the development and execution of SAIC’s NGA data production outsourcing programs. His teams repeatedly delivered high-quality standard and prototype products often under crisis conditions. Dr. Lobonc obtained a bachelor’s degree in Electrical Engineering and a Masters and PhD in Photogrammetry from Purdue University, studying under the esteemed Dr. Edward Mikhail. At Purdue he served as an instructor for numerous classes and was the lead researcher on various projects involving sensor modeling and automation of photogrammtric processes.

Nov. 16, Max J. Egenhofer, Prof. , University of Maine

Topological Reasoning on the Sphere

Popular models for reasoning about topological spatial relations, such as the 9-intersection, have been restrained to an embedding in the 2-dimensional plane. A new challenge comes from embedding the reasoning on the surface of a sphere. Is such spherical reasoning more complicated? Does it offer more possible inferences? Is it more precise? We will address these questions based on a systematic development of the binary topological relations that can be realized on the surface of a sphere. The study of the conceptual neighborhood graph of these spherical topological relations will reveal some stunning regularities. The subsequent analysis of the algebraic compositions of the spherical topological relations will provide measurable evidence for the inference power of spherical topological reasoning vs. planar topological reasoning. We will conclude with a discussion of how these results have an impact on the analysis of other scenarios of qualitative reasoning about spatial and temporal relations.

Max J. Egenhofer

National Center for Geographic Information and Analysis

Department of Spatial Information Science and Engineering

Department of Computer Science

University of Maine

Orono, ME 04469-5711, USA

max@spatial.maine.edu

http://www.spatial.maine.edu/~max/

Nov. 9, Carol Song, Ph.D, Rosen Center for Advanced Computing, Purdue University

Cyberinfrastructure for Environmental Sciences

Rosen Center for Advanced Computing at Purdue University is a research computing center that provides advanced computing and storage services to research scientists at Purdue and nationwide. We have embarked on a number of major cyber infrastructure projects in recent years. The TeraGrid project is an ambitious national undertaking to build and operate the world’s largest and most comprehensive grid computing cyberinfrastructure for open scientific research. RCAC/Purdue is one of the 9 partner universities/research institutions that provide computing, storage and data resources to the TeraGrid. As part of this effort, we have made available a collection of geospatial datasets, including the real-time satellite data (PTO), multi-spectral/hyper-spectral image data (LARS), and real-time Doppler radar data (NWS). We continue to integrate new datasets as they are demanded from the Purdue science community and work with users in utilizing these datasets. We recently started on a NSF-funded project, C4E4, in collaboration with CE/Agronomy/EAS/CS faculty to connect heterogeneous datasets and models that can be used to conduct scientific investigations on a range of real-world settings. These cyberinfrastructure efforts aim at connecting large communities of users through services for research, education, computing, visualization, and collaboration. This presentation will also demonstrate the capabilities of the Purdue Environmental Data Portal.

Dr. Carol Song is a Senior Research Scientist at the Rosen Center for Advanced Computing, Purdue University. She leads the Purdue TeraGrid Resource Partner project as the principal investigator and site lead. Dr. Song received her Ph.D. degree in computer science from the University of Illinois at Urbana-Champaign in 1992. She worked at NCSA (UIUC) for 5 years on scientific visualization. Dr. Song was a software architect and project manager at 3M and WAM!Net, developed commercial products for medical imaging and data center systems. She represented these corporations at the international standards working groups. Before joining Purdue in August 2005, Dr. Song led a startup company that develops innovative remote access software for mobile devices. She was awarded a National Science Foundation Small Business Innovative Research award and a research grant from the State of Indiana’s 21st Century Technology fund. Her current research interests include service-oriented architecture, mobile and pervasive computing software architecture, and grid computing middleware and tools and their applications in various scientific domains.

Nov. 2, Robert McMillan, CalTrans

Various geomatics projects executed by CalTrans, including lidar collection and processing done cooperatively with UC-davis, and the San Francisco Oakland Bay bridge construction project.

Oct. 26, Junhee Youn, Ph.D, Geomatics Engineering

Road extraction from true orthophoto and LIDAR

The primary subject of this research is an investigation into the automatic extraction of urban area roads from aerial imagery and LIDAR. Such extracted data becomes an essential component of a modern GIS system. In order to accomplish this, true orthophoto generated from aerial imagery and LIDAR height information is used to accurately fuse the data from multiple sensors and overcome the image displacements present in unrectified imagery. Next, the study area is subdivided or segmented based on homogeneity of the dominant road directions. Each region’s road candidates are selected with a proposed free passage measure. This process is called the Acupuncture method. Features around the road candidates are used as key factors for an advanced Acupuncture method named the Region Based Acupuncture method. For refining the road candidates, building and grass area thematic maps are generated from LIDAR and used as a blocking mask for roads.

Jun Hee Youn received his MS in Civil Engineering from Yonsei University, Korea. Also he was granted Ph.D degree in summer 2006 from the Purdue University, USA. His research interest is urban area feature extraction from the multiple sensors.

Oct. 19, Hank Theiss, Ph.D, Integrity Applications Incorporated (IAI)

Independent Validation of Sensor Models in the Community Sensor Model (CSM) Program

Students graduating from Purdue University's Geomatics Program have several career paths to consider. Three example choices touched on by this briefing include working directly for the Government, working as a tool/product developer, or working as contractor directly supporting the Government. In a scenario, to be discussed in the briefing, the Government hired private companies to develop sensor models to facilitate the exploitation of airborne imagery. In order to ensure that they were receiving sensor models that functioned photogrammetrically correctly, the Government hired a contractor (IAI) to develop a methodology to validate the implemented sensor models. The briefing outlines the series of tests performed, including both internal consistency checks and performance with real image and ground control data, and concludes with a software demonstration if time allows.

Dr. Hank Theiss has been the Chief Scientist in Photogrammetry for Integrity Applications Incorporated (IAI), a primarily Government Contractor company of 200+ employees in Northern Virginia in the suburbs of Washington DC since 2001. He is the team lead for six other photogrammetrists, all of whom completed Purdue's Geomatics Graduate Program and work on tasks that support the National Geospatial-Intelligence Agency (NGA). Hank completed his BS in Civil Engineering at Virginia Tech, under Professor Johnson, in 1994, went on to complete his MS and PhD degrees in Photogrammetry at Purdue University, with Professor Mikhail as Major Professor, in 2000, and stayed one year as a Visiting Assistant Professor.

Oct. 12, Charles Toth, Ph.D, The Center for Mapping, The Ohio State University

Mobile Mapping Supporting Autonomous Vehicle Navigation

The navigation of the robot vehicles in the GPS waypoints-defined corridors during the 2004 and 2005 DARPA Grand Challenges required the availability of a geographic database as well as a sophisticated sensor system to stay on track and avoid collisions. The latter sensing capability is practically identical to a state-of-the-art Mobile Mapping System’s ability, except in the autonomous vehicle’s case the real-time processing is a necessity. Mobile Mapping Systems are based on (1) direct orientation of the imaging sensors, which is typically achieved by GPS/INS integration, and (2) digital imaging sensors to sense and map the surroundings along the path of the vehicle. The environment for the upcoming 2007 Urban Challenges will be significantly different from the Grand Challenge races, which represented GPS waypoints-defined off-road vehicle navigation in open area, as besides GPS waypoints, route network description will be provided, including the stop sign locations, lane width, and checkpoint and parking locations. More importantly, this type of environment will require much more sophisticated sensing capabilities to cope with the object-rich urban surroundings where moving objects will also be present. In addition, achieving the required navigation performance will be more difficult in the GPS-denied/impeded urban landscape.

This presentation describes the mapping component of the OSU DARPA Challenge autonomous vehicle navigation system, as it transitions from the off-road to the urban environment. In particular, it provides an analysis of the overall mapping effort, including the development of a geospatial database, path-planning, interactive route analysis and the real-time mapping effort based on the experience of the OSU 2004 TerraMax and 2005 Desert Buckeye teams, followed by the new extensions, currently under research for the 2007 Urban Challenge race.

Dr. Charles Toth is a Senior Research Scientist at the Ohio State University Center for Mapping. He received an MS in Electrical Engineering and a Ph.D. in Electrical Engineering and Geoinformation Sciences from the Technical University of Budapest, Hungary. His research expertise covers broad areas of 2D/3D signal processing, spatial information systems, high-resolution imaging, surface extraction, modeling, integration and calibration of multi-sensor systems, multi-sensor geospatial data acquisition systems, and mobile mapping technology. He is chairing ISPRS (International Society of Photogrammetry and Remote Sensing) WG I/2 on LiDAR and InSAR Systems and serves as the Director for the Photogrammetric Application Division of ASPRS (American Society of Photogrammetry and Remote Sensing).

Oct. 5, Gilbert L. Rochon, Ph.D., MPH; Purdue Terrestrial Observatory; ITaP

Satellite Remote Sensing: The Interdisciplinary and International Dimensions

The history, development, proliferation and diversity of applications for earth observing satellite remote sensing are presented. The perspective offered is that civilian terrestrial remote sensing is a potentially equal opportunity technology, in that the same spatial, spectral and temporal resolutions are available for the developing countries as they are for the highly industrialized countries. Moreover, remote sensing's inherent multi-disciplinary utility combined with technological improvements allow remote sensing to be utilized in support of regional empowerment, disaster mitigation, public health and environmental sustainability. Real-time remote sensing, in combination with an accessible intelligent spatial data archive, enables communities to improve both preparedness for and responsiveness to an array of biogenic and anthropogenic disasters. Deployment of groundstations for real-time satellite data acquisition and dissemination at Purdue University are described as well as within the USA, Europe, Africa and Asia. Finally, the benefits of a global grid for archival and real-time remote sensing are delineated.

Gilbert L. Rochon

Associate Vice President for Collaborative Research

Director, Purdue Terrestrial Observatory

Chief Scientist, Rosen Center for Advanced Computing

Information Technology at Purdue (ITaP)

September 28, Larry Theller, Purdue Center for Advanced Applications in GIS

GIS Activities at the Purdue Center for Advanced Applications in GIS

This is a review of current projects that CAAGIS is involved with, and will include a discussion of the new horizons in GIS, as Indiana enters the age of streaming data services. We have supported the statewide efforts to provide data as streaming services, and those will be detailed. CAAGIS has several projects that extend the now-classic LTHIA online model into some interesting directions. Some of the proposals that have been advanced but are yet unfunded are also worthy of discussion.

Larry Theller began his career working as a geologist. After exploring the Western US, Canada, and the Gulf Coast for a few years he left fieldwork to work for several consulting firms in Denver, Colorado. He did computer-based geological projects and worked on developing early GIS and modeling software. Larry co-founded a consulting firm, " GeoPotential Inc." in Golden, Colorado where Larry headed a team of geologists, geophysicists and support personnel who developed oil and gas projects in the Rocky Mountains. When the western US oil and gas industry crashed his firm closed and he came to Purdue and earned a Master of Science Degree; He finished coursework as Purdue discussed forming a Center for Advanced Applications in GIS to promote GIS use across campus. CAAGIS was formed and Larry was hired by Bernie Engel to be the Geographic Information Specialist there. For the next 3 years CAAGIS used a Purdue reinvestment grant to buy the ArcView/Arc INFO license for the whole campus and to promote GIS in departments that were not using it. CAAGIS is essentially a "Virtual Center" with flexible associations to various faculty. It is driven by external funding ( grants ) and competes with several other Indiana organizations for money.

September 21, Mireille Boutin, Prof., School of Electrical and Computer Engineering, Department of Mathematics

Please notice the time and location different from the usual schedule:

2:00-3:00pm Thursday, Sep. 21, 2006, JNSN 245

Improving the Stability of Structure from Motion using Elimination Theory

Structure from motion (SFM) is the problem of reconstructing the geometry of a scene from a stream of images with tracked features. In this talk, we consider a projective camera model and assume that the internal parameters of the camera are known. Our goal is to reconstruct the geometry of the scene up to a rigid motion (i.e. Euclidean reconstruction.) It has been shown that estimating the pose of the camera from the images is an ill-conditioned problem, as variations in the camera orientation and camera position cannot be distinguished. Unfortunately, the camera pose parameters are an intrinsic part of current formulations of SFM. This leads to numerical instability in the reconstruction of the scene. Using algebraic methods, we obtain a new formulation of SFM which eliminates this cause of instability. In particular, this formulation can be used to formulate an improved bundle adjustment method which does not involve any camera angle. Numerical results demonstrate the improved robustness of this approach.

Mireille Boutin received the B.Sc. degree in Physics-Mathematics from the University de Montreal, and the Ph.D. degree in Mathematics from the University of Minnesota in Minneapolis. In 2001-2002, she was a visiting research scientist in the groups of David Mumford and of David Cooper at Brown University. In 2002-2003, she worked at the Max Planck Institute for Mathematics in the Sciences (Leipzig, Germany.) She joined Purdue’s School of Electrical and Computer Engineering in 2004 (with a courtesy appointment in the Department of Mathematics.). Her research interests include image and video processing, object recognition, automatic text translation, and computational mathematics. In 2002, with Kathryn Leonard, she co-founded the Rose-Whelan Society, an organization for women graduate students and post-doctorate in mathematics/applied mathematics at Brown University. She is a member of IEEE, the AMS and FoCM.

September 14, Laura Arns, Prof, Purdue Envision Center

Larry Biehl, Purdue Terrestrial Observatory

Research Resources and Activities at the

Purdue Envision Center and Purdue Terrestrial Observatory

This presentation will cover the resources that are available at the Purdue Envision Center, such as the Flex, Tiled Wall, Access Grid, haptics, and motion capture systems. Also presented will be several of the past and current projects at the Center, related to various disciplines, such as the immersive ecological experience, Art on the Grid, and sign language education. Capabilities of the newly installed tracking antenna at the Purdue Terrestrial Observatory to acquire MODIS, AVHRR and MVISR satellite data will be presented.

Laura Arns is the Associate Director and a Research Scientist for the Envision Center at Purdue University. She also holds a courtesy Assistant Professor appointment with the Purdue Department of Computer Graphics Technology, where she teaches a graduate course on virtual environments and an undergraduate course on computer graphics programming. Dr. Arns received her Ph.D. in Computer Science from Iowa State University in 2002, and also holds an M.S. in Computer Science from Iowa State University and a B.A. in Computer Science and Mathematics from Wartburg College. Her research interests are in the areas of applied virtual environments, human factors in virtual reality, and virtual reality usability.

Mr. Biehl is the Systems Manager for the Purdue Terrestrial Observatory at Purdue University. He received his Bachelor's Degree from Purdue University in Electrical Engineering (1973) and his Master's Degree from Purdue University in Engineering (1974). He was a research engineer with Purdue University's Laboratory for Application of Remote Sensing (LARS) from 1974 to 1985 and 1988 to 2004. He participated in Skylab and Landsat MSS and Thematic Mapper research, and had major responsibilities in NASA-sponsored field research programs including field spectral data acquisition and calibration procedures, data preprocessing and software development. During 1985 to 1988, he was the Operations Manager of the Cybotech Product Development Laboratory, a joint robotics project involving Purdue University engineering faculty and graduate students and Cybotech Corporation. Currently he is involved with the PTO, and IndianaView and is a resource person for remote sensing image data and application software including Leica Geosystems Imagine, RSI ENVI, Definiens Professional (eCognition) and MultiSpec.

September 7, Tonglin Zhang, Prof., Department of Statistics

Testing spatial autocorrelation in a loglinear model

Moran's I is the most widely used and the most frequently cited test statistic in spatial statistical literature. This research bridges the permutation test of Moran's I to the residuals of a loglinear model under the asymptotic normality assumption. It provides the versions of Moran's I based on Pearson residuals I_PR and deviance residuals I_DR so that they can be used to test for spatial clustering while at the same time account for potential covariates and heterogeneous population sizes. Our simulations showed that both I_PR and I_DR are effective to account for heterogeneous population sizes. The tests based on I_PR and I_DR are applied to a set of loglieanr models for early stage and late-stage breast cancer with socioeconomic and access-to-care data in Kentucky. The results showed that socioeconomic and access-to-care variables can sufficiently explain spatial clustering of early stage breast carcinomas, but these factors cannot explain that for the late-stage. For this reason, we used local spatial association terms and located four late-stage breast cancer clusters that could not be explained. The results also confirmed our expectation that a high screening level would be associated with a high incidence rate of early stage disease, which in turn would reduce late-stage incidence rates.

Tonglin Zhang got his PH. D. in statistics from the University of Michigan in 2002. After that he joined the Department of Statistics, Purdue University in fall 2002. Tonglin Zhang's research interest is mathematical statistics, spatial statistics for health and environmental data, and statistics problems in physics.

August 31, Ayman Habib, Prof., University of Calgary, Department of Geomatics Engineering, Canada

Multi-sensor and multi-primitive photogrammetric triangulation

The steady evolution in the mapping technology is leading to an increasing availability of multi-sensory geo-spatial datasets at a reasonable cost. For decades, analog frame cameras have been the traditional source of mapping data. The development of softcopy photogrammetric workstations together with the improved performance of CCD and CMOS arrays is stimulating the direct incorporation of digital imaging systems in mapping activities. However, current digital frame cameras are incapable of providing imagery while maintaining the geometric resolution and ground coverage of analog sensors. In spite of this deficiency, the low cost of medium-format digital frame imaging systems has lead to their frequent adoption by the mapping community. To offset the limitations of digital frame cameras, line cameras are being utilized by the majority of space borne imaging systems and some airborne platforms. In another front, the improved performance of the GPS/INS technology is having a positive impact in reducing the control requirement for photogrammetric triangulation. Besides such a contribution, the increased accuracy of direct geo-referencing systems is leading to a wide implementation of LIDAR (LIght Detection and Ranging) systems for 3D data collection. The complementary nature of the acquired spatial data by imaging and LIDAR systems are motivating their integration for complete description of the object space. However, such integration is only possible after accurate co-registration of collected data to a common reference frame. This presentation introduces algorithms for multi-primitive and multi-sensory triangulation environment, which is geared towards taking an advantage of the complementary characteristics of available spatial data from the above sensors. The triangulation procedure will ensure the alignment of involved data to a common reference frame as defined by the utilized control. The devised methodologies are tested and proved efficient through experimental results from multi-sensory real data.

Dr. Ayman F. Habib received an MSc in civil engineering from Cairo University, Egypt, an MSc and a PhD in Photogrammetry from the Ohio State University, USA. Currently, he is a professor at the Department of Geomatics Engineering, University of Calgary, Canada. His research interests span the fields of terrestrial and aerial mobile mapping systems, modeling the perspective geometry of non-traditional imaging scanners (e.g., line cameras), automatic matching and change detection between various datasets, automatic calibration of low cost digital cameras, incorporating analytical and free-form linear features in various photogrammetric orientation procedures, object recognition in imagery, and integrating photogrammetric data with other sensors/datasets (e.g., GPS/INS, GIS databases, multi- and hyper-spectral sensors, and LIDAR). Dr. Habib is recipients of several awards such as the Duane C. Brown Senior Award from The Ohio State University (1997), Talbert Abrams "Grand Award" from the ASPRS (2002), Talbert Abrams "Second Honorable Mention" from the ASPRS (2004), and the Research Excellence Award from the University of Calgary (2006). Dr. Habib has authored more than one hundred publications in peer reviewed journals, conference proceeding articles, technical reports, and lecture notes. Also his is one of the co-authors of a newly published book dealing with the acquisition, manipulation and applications of Digital Terrain Models.

August 24, Ali Cetin and Oguz Gungor, Geomatics Engineering

2006 Summer Intern Experience with ESRI, Redlands, CA

ESRI's Summer Internship Program can provide you with valuable work experience while you continue your education. Summer intern opportunities may include, but are not limited to, the following areas: Product Development, GIS Services, Internet Development, Educational Services, Accounting, Legal/Contracts, and Marketing. As an intern, we were involved in exciting team projects with some of the most talented people in the GIS software industry. Our internship was with the Release team in the Product Development team. In the seminar the steps for the application to the internship, our assignments, and the work environment will be presented.