Geographic Information Systems Spatial Data Models and their Implementation Komal Kapoor Ajit...

31
Geographic Geographic Information Systems Information Systems Spatial Data Models and their Spatial Data Models and their Implementation Implementation Komal Kapoor Ajit Padukone

Transcript of Geographic Information Systems Spatial Data Models and their Implementation Komal Kapoor Ajit...

Page 1: Geographic Information Systems Spatial Data Models and their Implementation Komal Kapoor Ajit Padukone.

Geographic Geographic Information SystemsInformation Systems

Spatial Data Models and their Spatial Data Models and their ImplementationImplementation

Komal KapoorAjit Padukone

Page 2: Geographic Information Systems Spatial Data Models and their Implementation Komal Kapoor Ajit Padukone.

AgendaAgenda• Spatial Data Models• Types of Spatial Data Models• Object Models

• Types• Implementation• Operations

• Field Models• Operations – Map Algebra• Raster Representation

• Surface Models• Raster-based• Vector-based

• GIS Libraries

Page 3: Geographic Information Systems Spatial Data Models and their Implementation Komal Kapoor Ajit Padukone.

Models - What are theyModels - What are they

• A model is an artificial construction in which parts of a source domain are represented in a target domain.

• The purpose of the model is to simplify and abstract away from the source domain.

source:  GIS: A Computing Perspective 2nd Edition, Worboys and Duckham

Page 4: Geographic Information Systems Spatial Data Models and their Implementation Komal Kapoor Ajit Padukone.

Geospatial Information Models - Geospatial Information Models - TypesTypes1) Object Models

2) Field Models

3) Surface Models

source:  Data Models in Commercial GIS Systems, Eric Hoel

Page 5: Geographic Information Systems Spatial Data Models and their Implementation Komal Kapoor Ajit Padukone.

1) Spaghetti Model

Simplest of the Object based data models.

Geometric Representations do not have any explicit relationships.

source:  Data Models in Commercial GIS Systems, Eric Hoel

Spaghetti Model of Hurricane Data

source: http://www.snowbirdnationals.com/weather.htm

Object Models - TypesObject Models - Types

Page 6: Geographic Information Systems Spatial Data Models and their Implementation Komal Kapoor Ajit Padukone.

2) Network Model

One-dimensional collection of interconnected Point (Junctions) and Line (Edges) features.    Facilitate modelling of constrained flow along edges and through junctions.

source:  Data Models in Commercial GIS Systems, Eric Hoel

Object Models - TypesObject Models - Types

Page 7: Geographic Information Systems Spatial Data Models and their Implementation Komal Kapoor Ajit Padukone.

A) Directed Network Models : Hydrological Networks

source: www.riversystems.washington.edu/story/Amazon

Object Models - TypesObject Models - Types

Page 8: Geographic Information Systems Spatial Data Models and their Implementation Komal Kapoor Ajit Padukone.

B) Undirected Network Models: Transport Networks

source: Assessing Spatial Aspects of School Location- Allocation in Copenhagen, Lasse Møller-Jensen

Object Models - TypesObject Models - Types

Page 9: Geographic Information Systems Spatial Data Models and their Implementation Komal Kapoor Ajit Padukone.

Object Models - TypesObject Models - Types3) Topological Model

Controls the Geometric relationships between features and maintain the geometric integrity.Represented in the form of Nodes, Arcs and Faces with explicit relationships between each other.

source:  Data Models in Commercial GIS Systems, Eric Hoel

source: http://spatialnews.geocomm.com/features/laserscan2/

Page 10: Geographic Information Systems Spatial Data Models and their Implementation Komal Kapoor Ajit Padukone.

The Open Geospatial Consortium, Inc.® (OGC)  leads the development of standards for geospatial and location based services.

The Simple Features Access Specifications define a standard Structured Query Language (SQL) schema that supports storage, retrieval, query and update of feature collections.

The OGC Specifications define a data model for the spatial data. The basic primitives are Point, Curve and Surface. Also are defined other useful structures like MultiPoint, MultiLine, MultiPolygon, etc

source:  OpenGIS Implementation Specification for Geographic information - Simple feature access

Object Models - Object Models - ImplementationImplementation

Page 11: Geographic Information Systems Spatial Data Models and their Implementation Komal Kapoor Ajit Padukone.

Object Models - Object Models - ImplementationImplementation

Geometry Class Hierarchy

source:  OpenGIS Implementation Specification for Geographic information - Simple feature access

Page 12: Geographic Information Systems Spatial Data Models and their Implementation Komal Kapoor Ajit Padukone.

OperationsSpaghetti Models

Network ModelsTopological

ModelsOGC Geometry

Model

equals Yes No Yes Yes

subset of Yes No Yes Yes

is empty Yes No Yes Yes

member of Yes No Yes Yes

disjoint from No No Yes Yes

intersection Yes No Yes Yes

union Yes No Yes Yes

difference No No Yes Yes

Set Oriented Operations

Object Models - OperationsObject Models - Operations

Page 13: Geographic Information Systems Spatial Data Models and their Implementation Komal Kapoor Ajit Padukone.

Object Models - OperationsObject Models - OperationsTopological Operations

OperationsSpaghetti Models

Network ModelsTopological

ModelsOGC Geometry

Model

boundary Yes No Yes Yes

interior No No Yes Yes

closure No No Yes Yes

meets Yes No Yes Yes

overlaps No No Yes Yes

is inside No No Yes Yes

covers No No Yes Yes

connected No No Yes Yes

components Yes No Yes Yes

extremes No No Yes Yes

is within No No Yes Yes

Page 14: Geographic Information Systems Spatial Data Models and their Implementation Komal Kapoor Ajit Padukone.

Object Models - OperationsObject Models - OperationsEuclidean Operations

OperationsSpaghetti Models

Network ModelsTopological

ModelsOGC Geometry

Model

distance No Yes Yes Yes

bearning/angle No No Yes Yes

length No No Yes Yes

area No No Yes yes

perimeter Yes No Yes Yes

centroid No No Yes Yes

Page 15: Geographic Information Systems Spatial Data Models and their Implementation Komal Kapoor Ajit Padukone.

• Represent spatial variations of an attribute over a set of

locations (spatial framework or support).

• Spatial framework comprise tessellations which can be

regular or irregular.

Regular tessellations Irregular tessellations : TIN

Field Based ModelsField Based Models

Souce : http://geosciences.tamuk.edu/~yu/web/L5312/lecture6.pdf

Page 16: Geographic Information Systems Spatial Data Models and their Implementation Komal Kapoor Ajit Padukone.

  TIN GRID

Advantages

•ability to describe the surface at different level of resolution•efficiency in storing data

•easy to store and manipulate•easy integration with raster databases

Disadvantages

•in many cases require visual inspection and manual control of the network

•inability to use various grid sizes to reflect areas of different complexity of relief.

ComparisonComparison of GRID and of GRID and TINTIN

Souce: http://www.ian-ko.com/resources/triangulated_irregular_network.htm

Page 17: Geographic Information Systems Spatial Data Models and their Implementation Komal Kapoor Ajit Padukone.

• Local o trigonometric, exponential, reclassification, selection, and statistical functionso E.g. mapping distances to categories ‘very close’, ‘close’, ‘far’, ‘very far’ (reclassification)o Syntax : sin(c:\data\inraster1)

• Focal o Neighborhood operations : majority, maximum, sum, range, slope calculation etc.o Types of Neighborhoods :o Syntax : focalsum([inlayer1], rectangle, 3, 3)

• Zonalo computing zonal statistics such as sum, mean, maximum etc.o Use for calculation of raster areaso Syntax : zonalmean([inlayer1], c:\spatial\inraster2)

• Globalo generation of Euclidean and cost distance metricso Syntax : eucdistance(e:\data2\inrastersource)

• Overlayso Uses multiple fields e.g. densityo Syntax: mean([inlayer1], [inlayer2], [inlayer3])

Map AlgebraMap Algebra

Souce: ArchGIS 9.2 Desktop Help

Page 18: Geographic Information Systems Spatial Data Models and their Implementation Komal Kapoor Ajit Padukone.

• Space is defined using equally sized cells arranged in rows and

columns

• Can comprise of single or multiple bands

Source: ArchGIS 9.2 Desktop Help

Each band represents a segment of the electromagnetic spectrum collected by a sensor.

Raster Data Raster Data RepresentationRepresentation

Page 19: Geographic Information Systems Spatial Data Models and their Implementation Komal Kapoor Ajit Padukone.

• MrSID

A compression technique especially for maintaining the quality of large images. Allows for a high-compression ratio and fast access to large amounts of data at any scale.

• ESRI GRID

A proprietary ESRI format that supports 32-bit integer and 32-bit floating-point raster grids.

• ERDAS Imagine

Produced using IMAGINE image processing software created by ERDAS. IMAGINE files can store both continuous and discrete, single-band and multiband data.

• TIFF(Tagged Image File Format)

Widespread use in the desktop publishing world. It serves as an interface to several scanners and graphic arts packages. TIFF supports black-and-white, grayscale, pseudo color, and true color images, all of which can be stored in a compressed or decompressed format.

Examples of RasterExamples of Raster

Souce: ArchGIS 9.2 Desktop Help

Page 20: Geographic Information Systems Spatial Data Models and their Implementation Komal Kapoor Ajit Padukone.

• Possible storage methods for multiple raster data:o store each raster dataset individuallyo mosaic them into one large raster dataseto store them as members of a raster catalog

• Raster catalogs can be used to display adjacent or overlapping raster's without having to mosaic them together into one large file.

• Each raster maintains it own properties: multiple formats, data types, file sizes, and coordinate systems.

Used for storing massive image repositories, retaining overlaps between datasets, managing time series data, and when differences among adjoining images prevent mosaicking.

Source: ArchGIS 9.2 Desktop Help

Raster CatalogsRaster Catalogs

Souce: ArchGIS 9.2 Desktop Help

Page 21: Geographic Information Systems Spatial Data Models and their Implementation Komal Kapoor Ajit Padukone.

• Used to represent the topography of the Earth

• Built using remote sensing data or survey data

• Some applications :oCreating relief mapsoRendering 3D visualizationsoModeling water flowoRectification of aerial photographyoTerrain analyses in geomorphology.

Surface Models Surface Models (Digital Evaluation Models(Digital Evaluation Models))

Souce: Data Models in Commercial GIS Systems, E. Hoel, page 215-219.

Page 22: Geographic Information Systems Spatial Data Models and their Implementation Komal Kapoor Ajit Padukone.

• Raster-based

• E.g. USGS DEM developed by the United States Geological Survey• Digital Terrain Elevation Data (DTED) developed by the National

Imagery and Mapping Agency (NIMA) in support of military applications

Source: http://www.esi.utexas.edu/gk12/workshops/gis/docs/elevation.ppt.

Souce: Data Models in Commercial GIS Systems, E. Hoel, page 215-219.

Surface Models (Contd…)Surface Models (Contd…)

Page 23: Geographic Information Systems Spatial Data Models and their Implementation Komal Kapoor Ajit Padukone.

• Vector-basedo TINs are used to partition the geographic space

• Advantages of TIN

• Facilitates a very close approximation of the surface

• Input vector data is incorporated directly in the model which can be queried as well.

• Preserves proximity based connectivity. Hence interpolation neighborhoods are always comprised of the closest input data/samples.

Source:http://www.geog.ucsb.edu/~good/papers/451.pdf

Surface Models (Contd…)Surface Models (Contd…)

Souce: Data Models in Commercial GIS Systems, E. Hoel, page 215-219.

Page 24: Geographic Information Systems Spatial Data Models and their Implementation Komal Kapoor Ajit Padukone.

Raster and Vector ModelsRaster and Vector Models

Inside Channel: Vector Outside Channel: Raster

3D Terrain Modelling

Source: http://www.ce.utexas.edu/stu/tateec

Page 25: Geographic Information Systems Spatial Data Models and their Implementation Komal Kapoor Ajit Padukone.

Raster and Vector ModelsRaster and Vector Models

Source: http://www.ce.utexas.edu/stu/tateec

Page 26: Geographic Information Systems Spatial Data Models and their Implementation Komal Kapoor Ajit Padukone.

GIS LibrariesGIS Libraries

Open Source GIS libraries providebasic functionality for certain aspects and tasks of GIS Software.

Most of them follow OGC's specifications regarding GIS Data Models and Operations.

Typically used for graphics, GIS format support and conversion, reprojection, computational geometry, topological operations.

Published under the MIT or LGPL licence. ( Can be used for commercial purposes without releasing source code. )

Source: Open Source GIS Libraries, Andreas Neumann, Marco Hugentobler

Page 27: Geographic Information Systems Spatial Data Models and their Implementation Komal Kapoor Ajit Padukone.

GDAL - Geospatial Data Abstraction Library        - library to access raster data formats.    - presents a single abstraction model       for all formats to an application    - written in C++    - Data Structures consist of a Dataset and a RasterBand.    - Dataset consists Georeferencing data.    - RasterBand consists the actual raster data.    - Used by :                                                                                        Google Earth                                    Grass GIS           

    

GIS LibrariesGIS Libraries

Source: Open Source GIS Libraries, Andreas Neumann, Marco Hugentobler

Page 28: Geographic Information Systems Spatial Data Models and their Implementation Komal Kapoor Ajit Padukone.

OGR

- part of the GDAL package.- deals with vector data formats like GML, ESRI shapefiles, GRASS and POSTgreSQL.- Data Structures include Layer, Feature and Geometry.- Feature contains the attribute values and reference to feature geometry.

GIS LibrariesGIS Libraries

Page 29: Geographic Information Systems Spatial Data Models and their Implementation Komal Kapoor Ajit Padukone.

JTS / JCS - Java Topological Suite / Java Conflation Suite.

- java libraries for 2D spatial functions.- supports the various spatial analysis methods and topological operations.- JCS is set of interactive tools for conflating and cleaning vector datasets.- used for coverage cleaning, coverage alignment and road network matching.- Projects using JTS : GeoServer, GeoTools, gvSIG, OpenJUMP, uDig.- Ported to C++ as GEOS- Ported to .NET as .NET Topology Suite.

GIS LibrariesGIS Libraries

Source: Open Source GIS Libraries, Andreas Neumann, Marco Hugentobler

Page 30: Geographic Information Systems Spatial Data Models and their Implementation Komal Kapoor Ajit Padukone.

GPSBabel

- cross-platform, free software to transfer routes, tracks, and waypoint data to and from consumer GPS units, and to convert between over a hundred types of GPS data formats.

- useful because GPS data has many proprietary formats.

- used by many collaborators to OpenStreetMap.

- also popular in "Geocaching" communities.

GIS LibrariesGIS Libraries

Source: Open Source GIS Libraries, Andreas Neumann, Marco Hugentobler

Page 31: Geographic Information Systems Spatial Data Models and their Implementation Komal Kapoor Ajit Padukone.

Questions?Questions?