Exploiting Imagery with ArcGIS
Exploiting Imagery with ArcGIS Brett Rose, Esri Ben Kamphaus, ITT-VIS
Workshop Outline
•
Imagery and ArcGIS
•
Exploitation -
Management, Visualization, Analysis & Dissemination
•
Advanced Exploitation with ENVI (ITT-VIS)
•
Summary
•
Q&A
ArcGIS 10 Provides the Imagery Platform
Dynamic Processing
Analysis & Extraction
Visualization & Analysis
Any Client
Image Service Supporting Time
Massive Scalability
ITT VIS BAE … Many Partners… …Many Opportunities
ArcGIS 10 Provides the Imagery Platform •
•
Providing imagery: -
Timely - Value is highest when new
-
Quickly - Fast display
-
Accurately - Correct location and metadata
Exploiting the rich information content: -
Resolution -
-
Temporal change -
-
Enable time control
Spectral range -
-
Use full spatial content with maximum quality
Utilize multiple bands
Dynamic range -
Utilize sensitivity for sensors
Image Management
The New Mosaic Dataset
•
Catalog/Library of all imagery & associated metadata -
Stored in geodatabase
-
References original pixels as files or database
-
Provides dynamic mosaicking and on-the-fly processing
-
Authored and accessible directly in ArcGIS Desktop -
•
Automation using geoprocessing tools & ArcObjects
Accessible as: -
-
Image -
Dynamically mosaiced
-
Processed on the fly
Catalog -
Table with geometry & metadata
Dynamic Mosaicing
•
Fuse of overlapping imagery from multiple sources
•
User control of Image Order -
By Date –’Latest’, ‘Closest to May 2001’
-
By Attribute – ‘Highest Sun Angle’
-
By Viewpoint – North, South, East, West
-
Seamline – Feathered blend
-
User Query – ‘Landsat imagery, with no cloud, later than June 2001’
On-Demand Processing
•
Define processing functions as part of Mosaic Dataset & imagery processed as accessed
•
Processes -
Orthorectify
-
PanSharpen
-
Extract (specify) Bands
-
Vegetation Index
-
Color Correction
-
Etc...
Mosaic Datasets
Storage
Processing Time
Reduces storage by removing redundancy
Reduces processing
Overlapping Imagery Maintain original information
Disparate Datasets Handle large NoData areas
t1 t2 t3 t4 t5
Multi-resolution Data No need to sample up or down
Maintenance Add imagery as required
Image Quality Reduces resampling
Maintain Metadata Valuable information
Image Visualization
Accelerated Display
•
Dynamic – Change Contrast, Brightness, Etc.
•
Accessible through Image Analysis Window
•
SW or HW based
•
Tile based access
The Image Analysis Window
•
Enhancement -
•
Interpretation -
•
Contrast, Brightness, Gamma, Dynamic Range Adjustment Sharpen, Top is Up
Processing Functions -
Clip, Mask, StackBands, NDVI, Orthorectify, PanSharpen
•
Has its own Layer List
•
Access to accelerated display
The Image Classification Toolbar
•
•
•
Image Classification Toolbar -
Supervised & Unsupervised
-
Class Probability
-
Principle Component Analysis
Training Sample Manager -
Define training areas graphically
-
Generate Signature
Signatures applied as functions -
Requires Spatial Analysis Extension
-
Not to be confused with Feature Extraction
Image Analysis
On-Demand Processing
•
Define processing functions as part of Mosaic Dataset & imagery processed as accessed
•
Processes -
Orthorectify
-
PanSharpen
-
Extract (specify) Bands
-
Vegetation Index
-
Color Correction
-
Etc...
On-Demand Processing
•
Traditional Image Processing
Original image Rectified image Multiple separate data products
•
On-Demand Processing
Multiple in-memory derived products (no additional storage requirements)
Original image Mosaic dataset (On-demand processing)
Raster Function
•
•
Functions added to each Mosaic Dataset -
Essentially algorithms that are applied dynamically when image is requested
-
Can create large chains
Example of Landsat 7 -
Compose a MS raster from 7 individual files
-
Create a Pan raster
-
Compose panshapened raster
-
Add fields to the catalog table Aquisition Date, Cloud Cover, etc
Image Dissemination
Three Patterns for Image Access
•
Direct (local file on disk)
•
Static (cache)
•
Dynamic / On-the-fly (local or web services)
•
ArcGIS Server (CORE) •
Directly from source raster dataset
•
Changeable Compression
•
Return rendered image or source pixel values
CORE
•
ArcGIS Server (Image Extension) • • • • • •
Directly from mosaic dataset Catalog (Search & Discovery) Dynamic mosaicing On-the-fly processing Changeable Compression Return rendered image or source pixel values SOAP REST Flex Silverlight JSAPI
Image Extension
Workshop Outline
•
Imagery and ArcGIS
•
Exploitation -
Management, Visualization, Analysis & Dissemination
•
Advanced Exploitation with ENVI (ITT-VIS)
•
Summary
•
Q&A
ArcGIS 10 Provides the Imagery Platform
Dynamic Processing
Analysis & Extraction
Visualization & Analysis
Any Client
Image Service Supporting Time
Massive Scalability
ITT VIS BAE … Many Partners… …Many Opportunities
ArcGIS 10 Provides the Imagery Platform •
•
Providing imagery: -
Timely - Value is highest when new
-
Quickly - Fast display
-
Accurately - Correct location and metadata
Exploiting the rich information content: -
Resolution -
-
Temporal change -
-
Enable time control
Spectral range -
-
Use full spatial content with maximum quality
Utilize multiple bands
Dynamic range -
Utilize sensitivity for sensors
Image Management
The New Mosaic Dataset
•
Catalog/Library of all imagery & associated metadata -
Stored in geodatabase
-
References original pixels as files or database
-
Provides dynamic mosaicking and on-the-fly processing
-
Authored and accessible directly in ArcGIS Desktop -
•
Automation using geoprocessing tools & ArcObjects
Accessible as: -
-
Image -
Dynamically mosaiced
-
Processed on the fly
Catalog -
Table with geometry & metadata
Dynamic Mosaicing
•
Fuse of overlapping imagery from multiple sources
•
User control of Image Order -
By Date –’Latest’, ‘Closest to May 2001’
-
By Attribute – ‘Highest Sun Angle’
-
By Viewpoint – North, South, East, West
-
Seamline – Feathered blend
-
User Query – ‘Landsat imagery, with no cloud, later than June 2001’
On-Demand Processing
•
Define processing functions as part of Mosaic Dataset & imagery processed as accessed
•
Processes -
Orthorectify
-
PanSharpen
-
Extract (specify) Bands
-
Vegetation Index
-
Color Correction
-
Etc...
Mosaic Datasets
Storage
Processing Time
Reduces storage by removing redundancy
Reduces processing
Overlapping Imagery Maintain original information
Disparate Datasets Handle large NoData areas
t1 t2 t3 t4 t5
Multi-resolution Data No need to sample up or down
Maintenance Add imagery as required
Image Quality Reduces resampling
Maintain Metadata Valuable information
Image Visualization
Accelerated Display
•
Dynamic – Change Contrast, Brightness, Etc.
•
Accessible through Image Analysis Window
•
SW or HW based
•
Tile based access
The Image Analysis Window
•
Enhancement -
•
Interpretation -
•
Contrast, Brightness, Gamma, Dynamic Range Adjustment Sharpen, Top is Up
Processing Functions -
Clip, Mask, StackBands, NDVI, Orthorectify, PanSharpen
•
Has its own Layer List
•
Access to accelerated display
The Image Classification Toolbar
•
•
•
Image Classification Toolbar -
Supervised & Unsupervised
-
Class Probability
-
Principle Component Analysis
Training Sample Manager -
Define training areas graphically
-
Generate Signature
Signatures applied as functions -
Requires Spatial Analysis Extension
-
Not to be confused with Feature Extraction
Image Analysis
On-Demand Processing
•
Define processing functions as part of Mosaic Dataset & imagery processed as accessed
•
Processes -
Orthorectify
-
PanSharpen
-
Extract (specify) Bands
-
Vegetation Index
-
Color Correction
-
Etc...
On-Demand Processing
•
Traditional Image Processing
Original image Rectified image Multiple separate data products
•
On-Demand Processing
Multiple in-memory derived products (no additional storage requirements)
Original image Mosaic dataset (On-demand processing)
Raster Function
•
•
Functions added to each Mosaic Dataset -
Essentially algorithms that are applied dynamically when image is requested
-
Can create large chains
Example of Landsat 7 -
Compose a MS raster from 7 individual files
-
Create a Pan raster
-
Compose panshapened raster
-
Add fields to the catalog table Aquisition Date, Cloud Cover, etc
Image Dissemination
Three Patterns for Image Access
•
Direct (local file on disk)
•
Static (cache)
•
Dynamic / On-the-fly (local or web services)
•
ArcGIS Server (CORE) •
Directly from source raster dataset
•
Changeable Compression
•
Return rendered image or source pixel values
CORE
•
ArcGIS Server (Image Extension) • • • • • •
Directly from mosaic dataset Catalog (Search & Discovery) Dynamic mosaicing On-the-fly processing Changeable Compression Return rendered image or source pixel values SOAP REST Flex Silverlight JSAPI
Image Extension
