Remote Sensing Systems, Remote Sensing Systems, Geographic Information Geographic Information Systems, and Systems, and the Classification of Urban the Classification of Urban Terrain Terrain Fred Cameron Fred Cameron Operational Research Advisor to Operational Research Advisor to Director General Land Combat Director General Land Combat Development Development Kingston, Ontario Kingston, Ontario
25
Embed
Remote Sensing Systems, Geographic Information Systems, and the Classification of Urban Terrain Fred Cameron Operational Research Advisor to Director General.
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
Remote Sensing Systems, Remote Sensing Systems, Geographic Information Systems, Geographic Information Systems,
andandthe Classification of Urban Terrainthe Classification of Urban Terrain
Fred CameronFred Cameron
Operational Research Advisor toOperational Research Advisor to
Director General Land Combat DevelopmentDirector General Land Combat Development
Kingston, OntarioKingston, Ontario
OutlineOutline
• Introduction and Historical Material– Ellefsen’s Study from 1980-86– Military Doctrine
• Sensors• Geographic Information Systems (GIS) and
Associated Analytical Tools• Queen’s University Study
– Geographical Information Systems and Remote Sensing– Metadata and Interoperability – DIGEST Standard– Artificial Intelligence and Rule Based Systems– Categorization, Land Cover, Land Use, and Semantics
• Models, Simulation, and Operational Research
Urban Terrain Zone ClassificationUrban Terrain Zone Classification
• Ellefsen’s Study, circa 1987– ‘Procedures’ and ‘Definitions’– ‘Urban Morphology’– ‘The Growth of Cities’ and ‘Structures and
Characteristic LiDAR IFSARFlight Altitude 2000 m above ground level 6000 m above ground level
Swath Width 540 mLevel III: 1600 mLevel IV: 630 m
Flight Speed 140 knots 180 knots
Collection Rate 25 square kilometers per hourLevel III: 50 sq. km per hourLevel IV: 25 sq. km per hour
ProcessingRate 3 hrs processing per 1 hr flight
Real-time onboardprocessing
Time Day or night Day or night
Weather No clouds, minimalprecipitation
No limitations
LiDAR Light Detection and RangingIFSAR Interferometric Synthetic Aperture Radar
Level I(Current Archive)
Level II(SRTM)
Level III(RTV)
Level IV(RTV)
Level V(RTV)
90 m spacing 30 m spacing 10 m spacing 3 m spacing 1 m spacing
Notional Difference in DTED ResolutionNotional Difference in DTED Resolution
DTED = Digital Terrain Elevation Data SRTM = Shuttle Radar Topographic Mission RTV = Rapid Terrain Visualization projectSource: US Army’s Rapid Terrain Visualization Project, Mr. Mike Hardaway, Technical Manager
Source: US Army’s Rapid Terrain Visualization Project, Mr. Mike Hardaway, Technical Manager
Assume: • first return is from top of tree canopy• last return is from the ‘ground’
Example: Line of Sight from LiDAR DataExample: Line of Sight from LiDAR Data
• ArcGIS Military Analyst methods applied to LiDAR data from Toronto
Source: Harrap and Lim, ‘Terrain Classification for Military Operations in Urban Areas’, 2003
Example: View Field from a PointExample: View Field from a Point
Field of view (green) from top of the Provincial Legislature in Toronto
Source: Harrap and Lim, ‘Terrain Classification for Military Operations in Urban Areas’, 2003
Example:Example:Building Extraction to GIS ShapesBuilding Extraction to GIS Shapes
• With some semantic assumptions, extraction of features can build GIS data with minimal intervention by an operator
• LIDAR Analyst, developed by Dr. Vincent Tao at York University, Toronto, does a good job on urban areas as shown.
Source: Harrap and Lim, ‘Terrain Classification for Military Operations in Urban Areas’, 2003
Pickering, Ontario
Bonn, Germany
Pan-chromatic Imagery
Classification by Alternate Methods
Classification by eCognition
Example from eCognitionExample from eCognition
Source: Birgit Mittelberg ‘Pixel Versus Object:A method comparison for analysing urban areas with VHR [very high resolution] data’ see http://www.definiens-imaging.com
Roles and UnderstandingRoles and Understanding
• Level of understanding is determined by process
• For Example (after Pigeon, 2002)– Sniper needs to have high spatial and environmental
texture resolution (i.e., the semantics of the immediate cover environment)
– Search and Rescue (SAR) pilot needs to have low spatial accuracy and high environmental texture resolution (i.e., the semantics of the landing zone environment)
– Blast models (physical) need medium to high spatial accuracy and accurate semantics of the target area
Modeling, Simulation, and OR AnalysisModeling, Simulation, and OR Analysis
• For Theoretical Analysis in Simulation: – Need representative terrain… but also– Need to know selected terrain is representative– Need to know ‘land use’ for entity behaviour
• For Rehearsal Analysis in Simulation:– Need actual terrain– Need to know ‘land use’ for entity behaviour
• For Mathematical Analysis:– Need terrain with appropriate characteristics– Do not necessarily need extensive raw data on
terrain, but need to know that assumptions in the model (sensor ranges, weapons ranges, lethal effects, etc.) are appropriate
MOUT FACT = Military Operations in Urban Environment Focus Area Collaborative Team
Models Covered by the Models Covered by the MOUT FACT AssessmentMOUT FACT Assessment
• Integrated Unit Simulation System (IUSS) – “constructive, force-on-force model, for assessing the combat worth of systems
and sub-systems for both individuals and small unit dismounted warfighters in high-resolution combat operations”
• CombatXXI
– “high-resolution, closed-form analysis tool for the assessment of new technologies”
– “replacement for CASTFOREM”
• AMSAA Infantry MOUT Simulation (AIMS)– “small unit combat simulation designed to support AMSAA systems
performance analyses of infantry systems”
• OneSAF Objective System– “composable, next generation computer-generated force (CGF) that can
represent a full range of operations, systems, and control processes from the individual combatant and platform level to battalion level”
• Mobility - Issues: NATO Reference Mobility Model V.2, decision-making on alternative paths through terrain
• Direct Fire - Issues: clearing buildings and hallways, deformable surfaces, non-lethal weapons, collateral damage, short-range engagements
• Wide Area Surveillance - Issues: radar, acoustics, SIGINT• Search and Target Acquisition - Issues: ACQUIRE model,
background noise, terrain and urban propagation, cues, shadows, rules of engagement, individual v. crew performance, and multiple targets
Source: Crino, ‘Representation of Urban Operations in Military Models and Simulations’
Model Assessment FindingsModel Assessment Findings
Source: Crino, ‘Representation of Urban Operations in Military Models and Simulations’
Needs ImprovementAdequate Poor
ConclusionsConclusions
• Dramatic remote sensing improvements for urban environments, e.g., LiDAR, IFSAR, multi-spectral and hyper-spectral cameras
• Rapid development in functionality of Geographic Information Systems, including imagery handling and automatic and semi-automatic classification
• Operational research practitioners need better understanding of cities and how they operate
• Coincidentally, so do military clients
ReferencesReferences• Scott T. Crino, ‘Representation of Urban Operations in Military Models and Simulations’ in
Proceedings of the 2001 Winter Simulation Conference, Dec 2001• Dispatches – “Training for Urban Operations”, Vol 9, No 2, Army Lessons Learned Centre,
Kingston, Ontario, May 2002• J-P Donnay, MJ Barnsley, and PA Longley, Remote Sensing and Urban Analysis, Taylor and
Francis, London and New York, 2001• Richard Ellefsen, Urban Terrain Zone Characteristics, US Army Human Engineering Lab,
Aberdeen, MD, 1987 • Rob Harrap and Kevin Lim, ‘Terrain Classification for Military Operations in Urban Areas’,
Queen’s University, Kingston, 2003• Jamison Jo Medby and Russell W. Glenn, Street Smart: Intelligence Preparation of the Battlefield
for Urban Operations, RAND, MR-1287-A, 2002• Bryan Mercer, ‘Comparing LIDAR and IFSAR: What can you expect?’ Proceedings of
Photogrammetric Week 2001• Birgit Mittelberg ‘Pixel Versus Object:A method comparison for analysing urban areas with VHR
[very high resolution] data’ Brochure from eCognition, see http://www.definiens-imaging.com• Luc Pigeon, ‘Concept of C4I data fusion command center for urban operations’ in Proceedings of
the 7th International Command and Control Research and Technology Symposium , Quebec, Sep 2002
• Jeffrey T. Turner and Christian P. Moscoso, ‘21st Century Terrain – Entering The Urban World’, Rapid Terrain Visualization Website: https://peoiewswebinfo.monmouth.army.mil/JPSD/rtv.htm , 2002