Subject: Satellite Imagery FAQ - 2/5
From: satfaq@pobox.com (Nick Kew)
Date: 17 Jan 1997 13:25:35 GMT
Archive-name: sci/Satellite-Imagery-FAQ/part2
This document is part the Satellite Imagery FAQ
Satellite Imagery
What are the main Earth Observation Satellites and Instruments?
------------------------------
Subject: Weather Satellites
Weather Satellites
_I know nothing about these: need to find some info._
The Meteosat GOES amd GMS weather satellites operate in geostationary
orbits. That is to say, they orbit the Earth at the same speed as the
Earth's rotation, thus constantly viewing the same area. This means
that their temporal resolution is effectively unlimited, so they are
able to generate the familiar weather 'movies'.
They are, however, of limited use for (other) remote sensing purposes.
Geostationary orbits (more typical of communications satellites) are
constrained to high altitude, and to the equator. Thus good viewing
angles over high latitudes are not possible. The very large area
images are at low spatial resolution; the best achieved by Meteosat
and GOES is 2.4Km (?).
Here are a few pointers to weather pictures online, or see the
Meteorology Resources FAQ for a far longer list.
------------------------------
Subject: Earth Observation Satellites (for geosciences, etc)
Earth Observation Satellites
_See also the list below, containing pointers to detailed information
and online imagery._
Earth Observation imagery takes a number of forms, of which the most
traditional are optical and near-infrared radiation, from about 0.4
(blue) to 2.0 (IR) micrometers. Examples include Landsat, Spot and
NOAA. These generally use tracking instrunents, the basic principles
of which are briefly described in Part 2 of this FAQ _(someone point
me to a proper intro on the net - SURELY there must be one)!_.
Colour
After basic processing, imagery from these satellites may appear as
photographs. With certain visual imagery - eg SPOT - it is even
possible to display images in more-or-less their natural colour. In
practice, images for display are generally manipulated to appear
visually pleasing and to show interesting detail, and appear in _false
colour_. Visible and non-visible (IR) bands may be freely mixed in
false colour images. There are no firm rules about this, but by
convention clouds are shown as white, and vegetation red or green,
depending on the context.
Resolution
Resolution is determined primarily by instrument design, and generally
involves various compromises:
1. High spatial resolution implies imaging a small area. For an image
of 1000 pixels square, at 20m resolution the area viewed is
20x20Km, but at 1Km resolution this increases to 1000x1000Km
(actually rather more, due to the variation in viewing angle over
a large area). The latter is therefore intrinsically suited to
large-scale studies.
2. High spatial resolution also implies a high sampling frequency,
which may limit the sensitivity of the sensor.
Types of Imagery
Apart from visual and near-infrared, other bands of the spectrum
commonly used include thermal infrared (heat) and microwave (radar).
Each of these has its own applications.
3-dimensional Imagery
We see the world in three dimensions by virtue of having two eyes,
viewing the world at slightly different angles. It is possible to
emulate this and produce 3-dimensional (stereo) satellite imagery, by
superimposing images of the same ground area, viewed from different
angles (and at different times). A limited number of satellites have
this capability.
------------------------------
Subject: Synthetic Aperture Radar (SAR)
Synthetic Aperture Radar
What is SAR?
Synthetic Aperture Radar. An active microwave instrument, producing
high-resolution imagery of the Earth's surface in all weather.
There is a good introduction to imaging radar by Tony Freeman of JPL
at http://southport.jpl.nasa.gov/desc/imagingradarv3.html
_Should we have an embedded intro for the benefit of non-WWW readers?
I can ask to include the above, or try and solicit an equally expert
intro from someone here_
What are the main SAR platforms?
Several past, present and future Earth Observation Satellites. Also
the Shuttle Imaging Radar missions. See the table for a full list.
* ERS-1/ERS-2
* JERS-1
* Shuttle Imaging Radar SIR-C/X-SAR
* Almaz
* RadarSat
the future...
* ENVISAT (I'm not even making a link until I've something REAL to
put there)!
* _OK, what have I forgotten about (or never heard of)?_
What distinguishes SAR from hi-res optical imagery?
Two main properties distinguish SAR from optical imagery:
* The SAR is an active instrument. That is to say, it generates its
own illumination of the scene to be viewed, in the manner of a
camera with flash. The satellite's illumination is coherent: i.e.
all the light in any flash is exactly in phase, in the manner of a
laser, so it does not simply disperse over the distance between
the satellite and the Earth's surface. A SAR instrument can
measure both intensity and phase of the reflected light, resulting
not only in a high sensitivity to texture, but also in some
three-dimensional capabilities. Experiments with the technique of
_Interferometry_ (measuring phase differences in exactly aligned
images of the same ground area) have shown that SAR can accurately
model relief, and appears able also to detect small changes over
time.
Some consequences of being an active instrument (and using
coherent light) are:
+ Works equally day or night
+ Polarised - can be used to gain additional information (esp.
when different polarisations are available on the same
platform - as on the most recent Shuttle missions).
+ Needs a lot more power than passive sensors, and can
therefore only operate intermittently.
+ Suffers from speckle, an artifact of interference patterns in
coherent light, sensitive to texture.
* SAR is _Radar_ - i.e. it uses microwave frequency radiation.
_(note that in consequence, references to "light" above should
more strictly read "microwave radiation")._ Microwave radiation
penetrates cloud and haze, so SAR views the Earth's surface (land
and sea) in all weather. For general purpose Remote Sensing, this
is probably _the_ major advantage of SAR.
An example of its use is the ESA/Eurimage "Earthwatch" programme,
producing imagery of natural and other disasters when weather
conditions prevent other forms of surveillence. Earthwatch imagery
is available at http://gds.esrin.esa.it/CSacquisitions
What are SAR images good for ?
* Sensitive to texture: good for vegetation studies.
* Ocean waves, winds, currents.
* Seismic Activity
* Moisture content
A list of SAR applications is available at
http://southport.jpl.nasa.gov/science/SAR_REFS.html
What is the meaning of colour in a SAR image?
Of course, all SAR image colour is false colour: the notion of true
colour is meaningless in the context of invisible microwave radiation.
Most SAR images are monochrome. However, multiple images of the same
scene taken at different times may be superimposed, to generate
false-colour multitemporal images. Colour in these images signifies
changes in the scene, which may arise due to a whole host of factors,
such as moisture content or crop growth on land, or wind and wave
conditions at sea. SAR is particularly well-suited to this technique,
due to the absence of cloud cover.
The shuttle SAR's images are the nearest to 'natural' colour, in the
sense that they are viewing three different wavelengths, which can be
mapped to RGB for pseudo-naturalistic display purposes (essentially
the same as false colour in optical/IR imagery).
_Need a proper multitemporal image entry_
_________________________________________________________________
Radar Altimetry
Technique used extensively to map the oceans. There are introductions
at http://www.satobsys.co.uk/ and http://dutlru8.lr.tudelft.nl/altim/.
The latter includes the _Altimetry Atlas_, computed from GEOSAT, ERS-1
and TOPEX-Poseidon altimetry data.
An interactive browser offering sea surface height maps is available
at http://www.ccar.colorado.edu/~hendricj/topexssh.html
_________________________________________________________________
------------------------------
Subject: List of some Earth Observation Satellites
What are the main Earth Observation Satellites and Sensors
_Here is a list of some EO missions. These entries should become html
links to further information (esp. details of imagery and where to get
it if applicable) on an ad-hoc basis, as and when I have the
information to put there (contributions sought) and the time to edit
them in._
For detail on any of the following (and others), try a keyword search
on Esrin's GDS at http://gds.esrin.esa.it/.
See also http://gds.esrin.esa.it/CIDN_PROVA.source
* ADEOS Advanced Earth Observing Satellite
+ OCTS Ocean Color and Temperature Scanner
+ AVNIR Advanced Visible and Near-Infrared Radiometer
+ NSCAT NASA Scatterometer
+ TOMS Total Ozone Mapping Spectrometer
+ POLDER Polarization and Directionality of the Earth's
Reflectance
+ IMG Interferometric Monitor for Greenhouse Gasses
+ ILAS Improved Limb Atmospheric Spectrometer
+ RIS Retroflector in Space
* Almaz
+ SAR
* DMSP Defense Meterological Satellite Program
+ SSM/I (Special Sensor Microwave/Imager)
+ Visible
+ SSM/T1, SSM/T2 Microwave temperature & moisture sounders
* ERS-1 Earth Resources Satellite
+ AMI (Active Microwave Instrument), Wind mode, Wave mode, SAR
(Synthetic Aperture Radar)
+ Radar Altimeter
+ ATSR-M (Along-Track Scanning Radiometer and Microwave
Sounder)
+ PRARE (Precise Range & Range Rate Equipment)
* ERS-2 as ERS1 with addition of
+ GOME Global Ozone Monitoring Experiment
* GEOS Geodynamics Experimental Ocean Satellite
* GEOSAT GEOdetic SATellite
* GMS Geostationary Meteorological Satellites (140 E)
+ VISSR (Visible and Infra-red Spin Scan Radiometer)
* GOES Geostationary Operational Environmental Satellite (75 W and
135 W)
+ VISSR (Visible and Infra-red Spin Scan Radiometer) altimeter
* HCMM Heat Capacity Mapping Mission
+ HCMR (Heat Capacity Mapping Radiometer), visible + thermal
* INSAT Geostationary satellite of India (74 E)
* IRS Indian Remote Sensing Satellite System
+ PAN - Panchromatic Camera
+ LISS I - III (Linear Imaging Self Scanning Sensors)
+ WIFS
* JERS-1 Japanese Earth Resources Satellite
+ OPS Optical Sensors
+ SAR (Synthetic Aperture Radar)
* KOSMOS Russian EO satellite
* Landsat
+ TM (Thematic Mapper)
+ MSS (Multi-Spectral Scanner System)
+ RBV (Return Beam Vidicon) camera
* METEOR Russian meteo satellites (2-21, 3-3, 3-5)
* Meteosat (0 E, Greenwich meridian)
+ Visible/near infra-red
+ middle IR
+ Watervapour, thermal infra-red
* MOS Marine Observation Satellite
+ MESSR Multispectral Electronic Self Scanning Radiometer
+ VTIR Visible and Thermal Infrared Radiometer
+ MSR Microwave Scanning Radiometer
* Nimbus 7
+ CZCS Coastal Zone Color Scanner
+ ERB Earth Radiation Budget
+ LIMS Limb Infra-red Monitor for the Stratosphere
+ SAM-II Stratospheric Aerosol measurement (II)
+ SAMS Stratospheric and Mesospheric Sounder
+ SBUV Solar and Backscatter ultraviolet Spectrometer
+ TOMS (Total Ozone Mapping Spectrometer)
+ SMMR (Scanning Multichannel Microwave Radiometer)
+ THIR Temperature Humidity Infra-red Radiometer
* NOAA Polar Orbiting Environmental Satellites (series)
+ AVHRR Advanced Very High Resolution Radiometer
+ TOVS (TIROS Operational Vertical Sounder)
+ SBUV/2 Solar Backscatter Ultraviolet Spectrometer
* Radarsat (Canada)
+ SAR
* RESURS
+ MSU-E High resolution optical scanner
+ MSU-SK Medium-resolution Optical-IR
* SeaStar
+ SeaWiFS Sea-viewing Wide Field-of-view Sensor
* SeaSat Ocean Dynamics Satellite
+ SAR L-band
+ ALT Radar altimeter
+ SASS Radar Scatterometer
+ SMMR Scanning Multi-Spectral Microwave Radiometer
+ VIRR Visible en Infra-red Radiometer
* Shuttle
+ SIR-A Shuttle Imaging Radar
+ SIR-B
+ SIR-C (cross polarized returns VH and HV) (Apr+Oct 1994)
+ LFC Large Format Camera
+ MOMS Modular Opto-electronic Multi-spectral Scanner (2 bands)
* SkyLab
+ S 192 MSS Multispectral Scanner
+ Metric camera experiment
* SPOT
+ HRV High Resolution Visible (2x) has 2 modes:
o XS (MultiSpectral mode)
o PAN (PANchromatic mode)
* SPOT 4 (launch 1995)
+ HRVIR High Resolution Visible and Infrared
* TIROS, TOS and ITOS forerunners of the current NOAA series
(9-12+14, 13 failed just after launch). See NOAA above.
+ AVHRR Advanced Very High Resolution Radiometer
+ TOVS (TIROS Operational Vertical Sounder) consisting of:
o HIRS/2 infra-red sounder
o SSU stratospheric sounding unit
o MSU microwave sounding unit
* TOPEX/POSEIDON
+ ALT Radar Altimeter
+ TMR TOPEX Microwave Radiometer
+ LRA Laser Retroreflector Array
+ SSALT Single-Frequency Solid-State Radar Altimeter
+ DORIS Dual-Doppler Tracking System Receiver
+ GPSDR GPS Demonstration Receiver
* TRMM Tropical Rainfall Measuring Mission (launch 1997, Japan)
+ PR Precipitation Radar
+ TMI TRMM Microwave Imager
+ VIRS Visible Infrared Scanner
+ CERES Clouds and the Earth's Radiant Energy System
+ LIS Lightning Imaging Sensor
_________________________________________________________________
------------------------------
Subject: Military / Intelligence Imagery
Military / Intelligence Imagery
FAS (Federation of American Scientists) have compiled a comprehensive
guide to imaging intelligence [IMINT] at
http://www.fas.org/irp/wwwimint.html.
Didn't President Clinton recently declassify some military imagery?
By an order dated 23rd Feb 1995,
* Imagery from the CORONA, ARGON, and LANYARD missions to be
declassified within 18 months.
* Review process to be instituted for other imagery.
Details and imagery are available at
http://edcwww.cr.usgs.gov/dclass/dclass.html.
_________________________________________________________________
------------------------------
Subject: Russian Imagery
Russian Imagery
_Contributed by W. Steven Sklaris (then of DBA systems; now
ssklaris@tds.com). Information regarding suppliers and availability
applies to the USA; elsewhere YMMV._
What about Russian Satellite Imagery?
The Russian Federation through the Russian Space Agency permits the
sale of commercial multi-source satellite imagery. The current
restriction placed on this imagery is limited to 2 meter resolution
but 1 meter resolutions are currently being considered. The majority
of commercial sources are from film return systems. The technical
philosophy is that the highest quality ground resolve is acquired by
film systems - no argument. The two primary commercial satellites are
KOSMOS, RESURS and Okean. The KOSMOS is utilized by the ministry of
Defense. RESURS and Okean satisfies environmental and weather
monitoring.
What are the characteristics of the KOSMOS satellite systems?
The KOSMOS has on board 2 camera systems; the KVR-1000 and TK-350. The
main attraction of the system is for mapping applications. The TK-350
is a frame camera that provides 80% overlap between images (every
third image provides 60%), along with internal and external
orientation data. This system provides for accurate determination of
latitude, longitude and elevation. The TK-350 covers an approximately
265 x 170 kilometer area per image and an 8 to 10 meter resolution.
The ground feature characteristics are provided by the KVR-1000
camera. This camera system operates simultaneously with the TK- 350
and provides 10% overlap between images. This is a panoramic camera
with 2 meter ground resolution and 36 - 44 x 165 kilometer area.
What are the characteristics of the RESURS satellite system?
The RESURS-O consists of the 01 and 02 series and are direct digital
return systems.
The RESURS-01 has on-board 2 sensor systems; the MSU-E and MSU-SK. The
MSU-E is a three channel system covering the 500 to 900 nanometer band
range. The sensor has a resolution of 45 meters and covers a 45
kilometer swath. The MSU-SK has 5 distinct channels covering the 540
to 11,800 nanometer band range. This sensor has a resolution of 160
meters for the first 4 channels and 600 meters for the 5th channel and
covers a 600 kilometer swath.
The RESURS-02 is an upgraded version of the 01 and has 4 on-board
sensor systems; the MSU-E, MSU-SK, SLR "Travers-1T" and MW-radiometer
"Delta-2." The MSU-E on this more recent satellite system covers the
same 3 channels as that of the 01 but the resolution has improved to
25 to 30 meters while retaining the 45 kilometer swath. The MSU-SK is
again included on the RESURS- 02 with no improvement from the 01
version. The Synthetic Aperture Radar "Travers-1T" and Micro Wave
radiometer "Delta-2" operate at a radiation wave length of 23cm. The
Travers-1T has a ground resolve of 200 x 200 meters and a swath width
of 100 kilometers. The Delta-2 has a ground resolve of 17,000 x 90,000
meters and a swath width of 1,000 kilometers.
The RESURS-F consists of the F1, F2 and F3 series.
The RESURS-F1 is the oldest and has on-board 2 camera types; the
KATE-200 and KFA-1000. The KATE-200 is a frame camera with a ground
resolution of 15 to 30 meters and covers a ground area of 240 x 240
kilometers. The camera system has three separate film bands covering
500 to 850 nanometers. The KFA- 1000 is an excellent higher resolution
color spectrazonal film camera and coverage of 80 x 80 kilometers. The
resolution advertised is 6 to 8 meters but is more around 8 to 10
meters. The color spectrazonal film covers the 570 to 680 nanometer
and 680 to 810 nanometer band ranges.
The RESURS-F2 is a more sophisiticated topographic camera system. The
MK-4 is a true multi-spectral camera system with data recorded on
three separate black and white film bases. There are 6 available bands
(460 to 900 nanometers) from which 3 can be selected for imaging. The
resolution of the MK-4 is about 6 to 8 meters and advertised to be
excellent for cartographic, environmental and geological surveys. The
coverage of the MK-4 is 150 x 150 kilometers. The RESURS-F2 has
on-board 2 stellar cameras to augment orientation accuracy information
but in almost all cases the cameras are not operated. Because of this
the cartographic capabilties are limited without ground control. The
excellent features of the camera are in the resolution and separate
band characteristics.
The RESURS-F3 is the most recent system and the most impressive. The
panchromatic frame camera covers 30 x 30 kilometers with at least 2
meter resolution. The 1:70,000 to 1:90,000 scale of the imagery
provides excellent ground definition.
What are the characteristics of the Okean?
The Okean-O is also a digital data return system and known to operate
for ocean monitoring. This satellite has on-board 6 sensor systems;
the MSU-V, MSU-SK, MSU-M, SLR, Scanning MW-radiometer "Delta-2", Track
MW-radiometer R- 600 and the Track VW-radiometer. The MSU-V is a eight
channel system, the spectral range is unknown. This sensor has a
resolution of 50 meters in the first 4 channels, 100 meters in the 6th
channel and 275 meters in the 7th and 8th channels and covers a 180 to
200 kilometer swath The MSU-SK has 5 distinct channels covering the
540 to 11,800 nanometer band range. This sensor has a resolution of
160 meters for the first 4 channels and 600 meters for the 5th channel
and covers a 600 kilometer swath. The MSU-M is a four channel system,
the spectral range is unknown. The sensor has a resolution of 1,600 to
2,000 meters and covers a 1,900 kilometer swath. The Side Looking
Radar operates at a radiation wavelength of 3.1cm at a ground
resolution of 800 to 1,500 meters and a swath width of 450 kilometers.
The Scanning Microwave radiometer "Delta-2" can operate at a
wavelength of 0.,8, 1.35, 2.2 or 4.5cm. The resolution is from 20,000
to 100,000 meters and covers a 800 kilometer swathwidth. The Track
Microwave radiometer R-600 operates at a wavelength of 6cm and has a
resolution of 130 meters (swath width unknown). The Track
VW-radiometer operates at a wavelength of 2.25 cm and also has a
resolution of 130 meters (swath width unknown).
What about all of the problems concerning Russian sources?
Numerous problems have been encountered with purchasing satellite
source from Russia. Most of the problems stemmed from the unauthorized
source distributors. Most distributors had access to the archives and
conducted a 1 or 2 time sale before they got caught. The Russian Space
Agency is now controling this distribution activity and has eliminated
this problem. Several other problems still exist and will not be
resolved in the near future. Access to coverage in a timely manner is
one. The archives of the KOSMOS system are not catalogued in a digital
form and acquiring coverage information is extensive and timely.
Information on coverage is typically provided in a week (depending on
the extent of coverage requested). The cloud cover information
provided with the coverage plots are very accurate but does not
satisfy all users. Several distributors of the TK-350 are preparing
digitization and browsing of the archived image files. Core Software
is considered to be the furthest along in this venture. A digital
database of the RESURS-F exists and provides extensive information
relating to coverage and collection detail. DBA Systems has a copy of
this database in their Melbourne, Florida office and can provide quick
turn-around information. The time to acquire the imagery has been
another problem area. This is much improved and is dependent on the
amount of coverage requested. A single image request, once selected
from the coverage plot, will take approximately 5 to 7 work days. Part
of this delay is due to the shipping services (DHL is 3 days from
Moscow). Film quality has also been questioned and although the
processing has significantly improved, many of the archived images are
scratched and were poorly processed during original production.
Can the film sources be provided in digital form?
Several distributors now provide the film sources in digital form.
EOSAT and DBA Systems both can perform digitization of the KVR-1000
down to the 45m range but only DBA can provide a continuous scan of
the entire TK-350 image down to the 45m spot size if desired (125m is
recommended). The precise scanning of their custom build scanner
retains the metric accuracy of the frame image. Any of the RESURS-F
films can also be scanned by the DBA scanner and JEBCO has also
provided digital product from the RESURS-F archives but we are unsure
whether the JEBCO source is still available. The color spectrazonal
film of the KFA-1000 cannot currently be captured by the DBA scanner
and other providers of color scanning of the KFA-1000 are unknown.
How can I purchase Russian Imagery?
There are several suppliers of Russian imagery and value-added
products created from the various Russian satellite systems. EOSAT,
through authorized Russian distributor Kieberso, provides digital
KVR-1000; Core Software through authorized Russian distributor
SOVINFORMSPUTNIK, provides hardcopy and digital KVR-1000 and TK-350;
DBA Systems through multiple authorized Russian distributors of
KOSMOS, RESURS and ALMAZ, provides the majority of Russian satellite
sources in both hardcopy and digital form.
Are the Russian planning any future commercial satellite systems?
Yes, the RESURS-F1M and RESURS-F2M will be upgrades to the existing
film return systems and a newer system referred to as Nika-Kuban will
be added to the RESURS satellite family. The Nika-Kuban will operate 3
camera systems and 1 forward looking digital return system to assist
in eliminating collection of cloud covered imagery. The Nika-Kuban
will offer panchromatic and multispectral collection in the 3 to 6
meter resolution range. Also planned as a major player in the
commercial remote sensing industry will be the ALMAZ-1B and ALMAZ-1C.
Both systems are currently awaiting financing to complete development
but will house the most sophisticated array of remote sensing systems
available in the commercial market. The ALMAZ-1B will offer a unique,
complex, multi-sensor payload providing for the first time, a
capability for simultaneous, multi-sensor, high resolution imagery,
including single-pass stereo coverage in the optical and multispectral
bandwidths; and high resolution, two-pass, all weather stereo in
microwave bandwidths.
Russian Imagery section by
W. Steven Sklaris
DBA Systems, Inc.
1200 South Woody Burke Rd.
Melbourne, Florida 32901
ph: 1-800-622-8554
fax: (407) 727-7019
------------------------------
Subject: Where can I get Imagery?
Where can I get Imagery?
This very frequently asked question has several parts, which are
addressed in various parts of this FAQ:
* Where can I get full products? (LIST - TBD)
* Where can I see/get samples of [some satellite's imagery] ?
* Where can I browse imagery for [some specific geographic
location]?
_Most of the references in this FAQ are global in scope - enter
lat/long or click a map.
_ * Where can I get current weather pics (online) ?
* Where can I browse images on the Web?
* Where can I get whole-world images?
* Where can I get full-resolution imagery cheap or free?
* Where can I get imagery for [my type of application]?
_That's outside the scope of this document - for the time being at
least - but check in the Further Reading_
------------------------------
Subject: How do I access the imagery catalogues?
How do I access the imagery catalogues?
There are a number of catalogue services available for interactive
login, via telnet; a few of these also offer alternative access
methods, including WWW. These will give full catalogue information,
and browse products online (typically by ftp). Some addresses for
these are listed under further reading.
CEOS IDN
The CEOS International Directory Network comprises three coordinating
nodes, together with a number of cooperating nodes. Each coordinating
node includes access to every known imagery catalogue, so in principle
you never need more than one address. These are listed in further
information.
Cintex
The Catalogue Interoperability Experiment aims to ensure
interoperability between the various catalogues.
GUIs for catalogue access
Various dedicated GUI systems exist to assist CINTEX catalogue users.
These include:
* DLR ISIS
* ESA UIT
* NASA EOSDIS V0 IMS
Details are available at http://gds.esrin.esa.it/Ccintex.cs.clients.
WWW Browse Services
In addition to the login services, there are some services available
on the WWW, offering a world-map and forms-based interface. These
include:
http://shark1.esrin.esa.it/
_Ionia_ AVHRR browser
http://tracy.esrin.esa.it:8001/
Eye-Browser Multi-Mission Browse Service: NOAA AVHRR, ERS-1
SAR, JERS OPS, Landsat TM.
http://www.coresw.com
"Imagenet" service - Landsat, SPOT and a promise of
Sovinformsputnik. Appears only to have data for America when
last checked. Commercial; the free service is limited.
http://www.eurimage.it/einet/einet_home.html
EiNet (European Imagenet) from Eurimage offers Landsat TM,
KVR-1000 and RESURS.
http://southport.jpl.nasa.gov/general.html/
SIR-C/X-SAR (Space Shuttle) imagery.
http://ic-www.arc.nasa.gov/ic/projects/bayes-group/Atlas/Earth/
Browser for Earth Observations from Shuttle
------------------------------
Subject: Where can I get full-resolution imagery cheap or free?
Where can I get full-resolution imagery cheap or free?
Answer 1: In general, you can't!
Answer 2: Old Landsat. The following was posted by Wim Bakker on
IMAGRS-L:
Paul DeVries (bosse@bahnhof.se) writes:
> Can anyone point me in the direction of satellite imagery of (dry) Andean
> altiplano, very cheap or in the public domain, of any vintage? Thanks.
In principle the old Landsat TM (acquired from July 16, 1982 through
September 27, 1985) and old Landsat MSS (older than 2 year) are
available at reduced prices:
MSS $ 200
TM raw $ 300
TM systematic corrected $ 425
TM precision corrected $ 600
Inquires can be made to
Customer Services
EROS Data Center (EDC)
Sioux Falls SD 57198
(605)-594-6151
In the mean time you can check on the Inventory service of EDC
URL telnet://glis.cr.usgs.gov
whether any images of your area of interest are available.
What datasets are available on CD-ROM?
Wim Bakker's report "Remote Sensing Data and GIS data on CD-ROM" is
available at http://www.itc.nl/~bakker/info/rs-data/index.html
Note - this is referenced for want of a better list, but is not kept
up-to-date.
_________________________________________________________________
------------------------------
Subject: Whole-World Images
Whole-World Images
_This answer is slanted towards Global AVHRR Land datasets. Anyone
care to talk about other images?_
Why create whole-world images?
_Because they're fun, of course! :-)_
Continental to global scale images are useful if they show
information that is studied at a large scale, such as the state
of the global biosphere. One major measure is NDVI, which
characterises 'greenness' (see RS/Vegetation FAQ for details).
Global NDVI images taken regularly over time - at intervals
between one and two weeks - enable scientists to study change
in the biosphere in detail.
How do they create whole-world images
The AVHRR Pathfinder and Global 1KM projects have created
global land datasets showing NDVI (together with lower-level
data) from AVHRR imagery, at resolutions up to 1.1KM. The
global images are created by mosaicing a large number of
individual scenes, taken over ten-day periods. Individual
scenes are first stitched to generate half-orbits (in principle
south to north pole, but generally broken because only daytime
data is used)! The half orbits are then stitched together, with
reference to a digital chart of the world.
The key to compositing for NDVI is that each point on the
Earth's surface is replicated in several images over the
sampling period. Only the _best_ NDVI value is selected, so bad
data (such as cloud cover) is discarded.
Why AVHRR? Why not, say, Landsat?
Yes, Landsat data is just as well-suited to computing NDVI as
is the AVHRR.
The NOAA satellites, in a polar orbit at an altitude of 833 KM,
orbit the Earth fourteen times per day. The AVHRR instrument
images a 2400-KM wide swath as it passes. Thus every point on
the Earth's surface is viewed at least about once per day (the
exact frequency of course varies with latitude).
The Landsat series (4-5), in near-polar orbits at 705 KM, also
orbit the Earth fourteen times per day. However, the swath
imaged is just 185KM, so a point on the equator may be viewed
only once in sixteen days. The data with which to generate
weekly, ten-day or fortnightly global composites is simply not
available. A sixteen-day composite would of course be subject
to considerable cloud-cover (see below).
Having said that, it is certainly possible to make large-area
Landsat mosaics. NASA's Landsat Pathfinder Project (see
http://pathfinder-www.sr.unh.edu/pathfinder/) has created such
datasets for the study of tropical deforestation.
How do they get rid of the cloud?
As noted above, only the best NDVI values from each input
dataset is used. Clouds will necessarily generate very low NDVI
values - _clouds are not green!_. Hence clouds are
automatically filtered out in the compositing process, provided
there is at least one cloudless view of a point within the
sample. Thus cloudlessness is not in fact guaranteed, but is
statistically far more likely than for a single pass.
Alternatively, it can be assured by collecting data over an
unlimited time period; c.f. the GeoSphere project).
Clearly this will work if and only if the characteristics being
studied are dissimilar to any cloud in at least one of the
available bands!
Further reading:
http://sun1.cr.usgs.gov/landdaac/1KM/1kmhomepage.html
Global Land 1-KM Project Front Page from USGS/EDC. Includes
extensive description of the project, and access to the data.
http://atlas.esrin.esa.it:8000/
Global AVHRR 1KM Server from ESA/ESRIN. The contents is
essentially the same as the EDC server; readers should normally
use whichever is closer to you in terms of Net connections.
http://shark1.esrin.esa.it/
_Ionia_ browser - AVHRR scenes and a browse version of a global
composite from ESA/ESRIN
http://xtreme.gsfc.nasa.gov/
AVHRR Land Pathfinder from NASA/GSFC - various global
composites.
http://infolane.com/infolane/geosphere/geospher.html
The GeoSphere project (commercial)
All the above references deal with global land datasets. NASA's
pathfinder program created also Ocean and Atmospheric datasets:
http://sst-www.jpl.nasa.gov/
SST Pathfinder from NASA/JPL
http://pegasus.nesdis.noaa.gov/pathfinder.html
Atmosphere pathfinder from NOAA
General Questions
------------------------------
Subject: Programmes and Policies
What are the National and International Remote Sensing programmes around the
World?
_(Should I have a brief summary and/or plain list here?_
This is dealt with in detail in a US Congress (Office of Technology
Assessment) report "Remotely Sensed Data: Technology, Management and
Markets", Chapter 5. Whilst this *is* explicitly a US government
document, it is generally an objective summary!
The report is available online at http://otabbs.ota.gov/T90 (thanks to
Mark_Goodman@achre.gov for drawing my attention to the OTA reports).
Where can I read about government policies in Remote Sensing
_USA_: See also the previous question.
The US Congress (Office of Technology Assessment) has published some
detailed reports, two of which are available online. In addition to
the report referenced in the previous question above, "Civilian
Satellite Remote Sensing: A Strategic Approach" is available at
http://otabbs.ota.gov/T85.
_Others_: AFAIK no such government documents are available elsewhere
(but see CEOS below for worldwide policy coordination). Check the
various space agency pages, listed under URLS.
------------------------------
Subject: Where can I find information on RS and the Environment
Resources concerning the Environment
This is far too big a subject to cover in this FAQ, so here are some
links, limited to major (and established) collections:
Environmental Resources Information Network, ERIN (Australia)
The ERIN homepage is at http://kaos.erin.gov.au/erin.html
(formerly listed under misc. URLS)
Global Environmental Research Federal Metadata Network GENIE
at http://www-genie.lut.ac.uk/.
United States Geological Survey - Environment
http://www.usgs.gov/environment/index.html
United Nations Environment Programme
Frontpage is at href=http://www.unep.ch. The main RS/GIS
related information is in the Global Resource Information
Database (GRID) at sites including
http://www.grid.unep.ch/gridhome.html, http://www.grida.no/ and
http://www.inpe.br/grid/home
US Global Change Research Information Office (GCRIO)
http://www.gcrio.org/
------------------------------
Subject: Using imagery during Natural (and other) disasters.
Can satellite imagery be used to watch newsworthy events?
Earthquakes, floods, volcanos, mega-icebergs, pollution disasters...
There is imagery for all of them! Watch relevant newsgroups as news of
a disaster breaks.
That's not to say there is immediate and extensive coverage of every
possible event: the satellites capable of imaging it may not be in the
right place at the right time! However, systematic programmes exist;
notably the ESA/Eurimage Earthwatch program at
http://www.eurimage.it/Earth_Watching/Earth_Watching.html _(formerly
listed at http://gds.esrin.esa.it/CSacquisitions which is still valid)_
------------------------------
Subject: Jobs
Where can I advertise or look for a job in Remote Sensing?
_Note: there is a very high percentage of duplication between these
sources!_
* The University of Minnesota's _GIS Jobs Clearinghouse_ at
http://www.gis.umn.edu/rsgisinfo/jobs.html. A good one-stop shop,
with the best list of pointers to other sources you'll find
anywhere.
* The GIS-JOBS list at gopher://nisp.ncl.ac.uk:70/11/lists/gis-jobs
* SPIE's Employment Service, at
http://www.spie.org/web/employment/employ_home.html
* The GEOSCI-JOBS and MET-JOBS listserv. Send subscription requests
(for both lists) to listproc@eskimo.com. You will recieve details
on how to post to the list, and guidelines for what is
appropriate. Either full (each job mailed separately) or digest
(weekly list) forms are available:
subscribe geosci-jobs-digest / met-jobs-digest (digest)
or subscribe geosci-jobs / met-jobs (full)
* Geographic Designs, Inc, are an agency specialising in RS/GIS.
http://www.geodesigns.com/
* GeoSearch, Inc are at http://www.geosearch.com/
* The GeoWeb Jobs Page http://www.ggrweb.com/job.html.
* SDCSC Jobs Page
In addition to the above, comp.infosystems.gis tolerates a certain
range of job postings. Please read the detailed guidelines in that
group's FAQ before posting.
_________________________________________________________________
------------------------------
Subject: Online Services Exchanges / Trade Fairs
WWW Information and Services Exchanges
The following interactive web sites are perhaps best described as
'trade fairs':
* European Wide Service Exchange http://ewse.ceo.org/
* GeoWeb http://www.ggrweb.com/
A similar but non-interactive site is
* The Geo Exchange http://giant.mindlink.net/geo_exchange
------------------------------
Subject: Geoscience Journal Information
Geoscience Journal Information
The UCSD service referenced in the August96 update of the SATFAQ drew
quite a lot of error reports, and has been withdrawn from here.
Elsevier have a mail server offering the tables of contents of their
Earth and Planetary Science journals. The subscription address for all
titles is earth-e@elsevier.nl. For information on the service, use
subject line "help".
A good reference point on the Web is Bill Corner's site, at
http://www.man.ac.uk/Arts/geography/rs/rs_journal.html.
------------------------------
Subject: Software + hardware
Software + hardware
Here's a complete cop-out: software is rather well covered in related
documents.
Where can I find Descriptions/Reviews of Remote Sensing Software?
There is an excellect collection of reviews, now maintained by Vinton
Valentine at
http://triton.cms.udel.edu:80/~oliver/gis_gip/gis_gip_list.html. In
spite of the "gislist" name, this deals extensively with Remote
Sensing and Image Processing software. Furthermore, comments and
reviews are generally independent of the manufacturers/distributors.
Is there a list of Software Vendors?
Where can I find information on Software Packages?
These questions are covered in the comp.infosystems.gis FAQ and the
"Using the Web for Geoscience Resources" FAQ, among others.
What software is available in the Public Domain?
See the Public Domain Cartographic Software FAQ.
Pointers to the FAQs are here.
Free packages for image processing include:
* Khoros, from ftp://ftp.khoros.unm.edu/ /
http://www.khoros.unm.edu/.
There is also a commercial khoros from khoral.com (frontpage
www.khoral.com)
* Grass, from ftp://moon.cecer.army.mil/
* MultiSpec from http://dynamo.ecn.purdue.edu/~biehl/MultiSpec/
A few more listed FYI with no comment (in all but one case, simply
because I know nothing):
* http://dcz.gso.uri.edu/XBrowse/browse/browse.html XBrowse- A
client-server browse application for satellite AVHRR imagery.
* Land Analysis System, from USGS/EDC (Landsat TM & NOAA AVHRR)
* http://www.atmos.washington.edu/gcg/SV.man/SVmanual.html Satview
(University of Washington).
How can I recieve imagery on my PC?
This question is dealt with in detail in the WXSAT FAQ and other
documents at ftp://kestrel.umd.edu/pub/wxsat/docs/.
There is a nice "Build your own HRPT groundstation" webpage at
http://www.msoft.it/noaa95/.
------------------------------
Subject: Standards
Standards Committee
Committee on Earth Observations Satellites (CEOS)
_I hope reproducing this paragraph isn't violating copyright - anyone?
It comes from too many sources to attribute!_
CEOS was created in 1984 as a result of the international Economic
Summit of Industrialized Nations and serves as the focal point for
international coordination of space-related, Earth observation
activities. Policy and technical issues of common interest related to
the whole spectrum of Earth observation satellite missions and data
received from such are addressed. CEOS encourages complementarity and
compatibility among space-borne Earth observing systems through
coordination in mission planning, promotion of full and
non-discriminatory data access, setting of data product standards, and
development of compatible data products, services, and applications.
The user community benefits directly from this international
coordination.
The CEOS information system is at http://gds.esrin.esa.it/CCEOSinfo,
and contains full details and CEOS files.
See also CEOS calibration pages at
http://southport.jpl.nasa.gov/calceos/calceos.html
CEOS also sponsors
The CEOS International Directory Network (CEOS IDN)
_Need someone to wirte a real entry_ This is the authoritative
worldwide information system that answers every possible question
about Satellite Earth Observation. The complete database is held at
the three coordinating nodes in America (NASA/GSFC), Europe
(ESA/ESRIN) and Asia (NASDA/EOC). For access details, see under
Further Information.
------------------------------
Subject: Copyright
How does Copyright affect Satellite Imagery?
Wim Bakker recently supplied the following article, in part a
translation from a (Dutch) NLR article. I have taken the liberty of
cutting it down somewhat.
I understand the issue of copyright on satellite imagery may in fact
vary significantly depending on what country you're in. Mark Goodman
(Mark_Goodman@achre.gov) writing from a US point of view comments:
I'm not sure that satellite imagery is covered by copyright law. It
may depend on what country you're in. I believe that SPOT and EOSAT
protect their intellectual property rights through trade secrets
laws, and through restrictive sales contracts that prohibit
redistribution of raw data, even for scientific use!
Your mileage may vary!
) Copyright
There is a lot of confusion about the copyright connected to the use
of satellite images and everything related to this.
According to Websters dictionary "copyright" is
1. copy.right \-.r{i-}t\ n : the exclusive legal right to reproduce,
publish, and sell the matter and form of a literary, musical, or
artistic work - copyright aj
2. copyright vt : to secure a copyright on
In 1886, during the Convention of Bern the matter of copyright was
regulated internationally. It states that the author (creator) of a
certain matter remains the owner of his product. This also means that
if you buy a copyrighted product you pay for the _use_ of this product
and you can never claim to be the owner of such a product.
Furthermore, you can never claim any other rights about such a product
(e.g. the right to _reproduce_ the product).
In copyright the following 5 stages can be distinguished:
1. the _creation_ of a product
2. the _manufacturing_ of a product
3. the _distribution_ of a product
4. the _use_ of a product
5. the _reproduction_ of a product
These 5 points can also be distinguished with the use of satellite
images. Two operational Earth observing satellites will be described
here: Landsat and SPOT.
_Here I have cut a detailed description of Landsat and SPOT
distribution, as being (IMHO) too detailed for this FAQ - NK._
Now when does the copyright principle touch the user?
Only when the user reproduces or copies (point 5) the satellite images
is he affected by the copyright issue. At all times the user must be
aware of the owner/producer of the data. The owner/producer may or may
not permit the reproduction of the datas, but must in any case be
mentioned on all publications of satellite images!
_Note: the following details may vary in different parts of the world,
although the principles apply in any case._
For SPOT data this will be CNES; for Landsat data received by European
ground stations this will be ESA; and for Landsat data from America
this will be EOSAT (or NOAA and EROS Data Center (EDC) for old data).
The owner/producer indicates which reproductions are allowed. The
reproduction of raw data - copying CCT's and film - is _never_ allowed
and for other categories that are allowed the owner will ask for a
certain contribution for the right to reproduce the data; this is
called the _reproduction fee_.
The following reproductions are free of reproduction fee
* Posters, slides, advertisement or publications used for
conferences, meetings, symposiums and exhibitions in the field of
Remote Sensing.
* Technical reports of RS conferences, symposiums etc.
* Scientific reports and papers
For the following, a reproduction fee is due:
* Newspapers
* Magazines
* Brochures
* Books _not_ related to the field of RS
* Posters, either ones that are sold as well as free copies
* Calendars
* Atlasses
* Postcards and invitations
* Using images on TV and video
At all times the owner/producer must be mentioned on the
reproductions, even if no reproduction fee is due!
This can be done in two ways
1. To use the word _copyright_ followed by the owner/producer and the
year of production. E.g.
Copyright ESA 1988
2. To use the international sign for copyright _)_ followed by the
owner/producer and the year of production. E.g.
) CNES/NLR 1994
In the last example the NLR could have processed data from SPOT.
Conclusion
* For some (scientific) applications you owe no _reproduction fee_.
* At all times the owner/producer must be mentioned on reproductions
using the word _copyright_ or the sign _)_
* In case of doubt, ask your distributor!
Subject: Satellite Imagery FAQ - 3/5
From: satfaq@pobox.com (Nick Kew)
Date: 17 Jan 1997 13:25:41 GMT
Archive-name: sci/Satellite-Imagery-FAQ/part3
This document is part of the Satellite Imagery FAQ
------------------------------
Subject: Image Basics
Image Basics _Contributed by Wim Bakker (bakker@itc.nl)_
What is an image?
A digital image is a collection of digital samples.
The real world scene is measured at regular distances (=digital). One
such measurement is limited in
* Space
One sample covers only a very small area from the real scene.
* Time
The sensor needs some integration time for one measurement (which
is usually very short).
* Spectral coverage
The sensor is only sensitive for a certain spectral range.
Furthermore, the sample is quantized, which means that the physical
measure in the real world scene is represented by a limited number of
levels only. Usually 256 levels of "grey" are sufficient for digital
images; 256 levels can be represented by an eight bit unsigned Digital
Number (DN). "Unsigned" because the amount of light is always
positive. More levels will need more bits; the quantization determines
the amount of bits per pixel on the image storage.
Image samples are usually called _pixel_ or _pel_ after the
combination of "picture" and "element". A pixel is the smallest unit
of a digital image. The size of this unit determines the resolution of
an image. The term _resolution_ is used for the detail that can be
represented by a digital image. As discussed before the resolution is
limited in four ways:
------------------------------
Subject: Resolution
* Spatial resolution.
If one pixel is a ground cell sample of 20 by 20 meter then no
objects smaller than 20 meter can be distinguished from their
background. This doesn't necessarily mean they cannot be
_detected_!
Note that if the spatial resolution doubles, the amount of image
data increases by a factor 4!
* Temporal resolution.
A distinction can be made between
+ Temporal resolution of one image.
Fast moving objects will appear blurred on one image. E.g.
the temporal resolution of one TV image is about 1/25 of a
second.
+ Temporal resolution of a time series of images.
If the images are taken sparsely in time then the possibility
exists that some phenomena will be missed. The resolution of
Landsat is 16 days, of SPOT 26 days and of NOAA 4 hours. So
the latter satellite is said to have a _high_ temporal
resolution even though the spatial resolution is _low
_compared to the two other satellites! (1.1 km and 20-30 m)
* Spectral resolution.
Current imaging satellites usually have a broad band spectral
response. Some airborne spectrometers exist that have a high
spectral resolution; AVIRIS Airborne Visible/Infrared Imaging
Spectrometer (from NASA/JPL) has 224 bands, GERIS Geophysical and
Environmental Research Imaging Spectrometer has 63 bands.
* Quantization.
E.g. if 100 Lux light gives DN 200 and 110 Lux yields DN 201 then
two samples from the original scene having 101 and 108 Lux will
both get the DN 200. Values from the range 100 up to 110 Lux can
not be distinguished.
======================== Image Formats (HTML) ======================
_Contributed by Wim Bakker (bakker@itc.nl)_
------------------------------
Subject: Image Formats
Image data on tape
Looking at the images stored on tape there's three types of
information
* Volume Directory, which is actually meta-information about the way
the headers/trailers and image data itself are stored
* Information about the images
This information can be stored in separate files or together with
the image data in one file.
This information can be virtually anything related to the image
data
+ Dimensions. Number of lines, pixels per line and bands etc.
+ Calibration data
+ Earth location data
+ Orbital elements from the satellite
+ Sun elevation and azimuth angle
+ Annotation text
+ Color Lookup tables
+ Histograms
+ Etc. etc...
The information is often called a _header_, information _after_
the image data is called a _trailer_
* The pure image data itself
The image data can be arranged inside the files in many ways. Most
common ones are
* BIP, Band Interleaved by Pixel
* BIL, Band Interleaved by Line
* BSQ, Band SeQuential
If the pixels of the bands A, B, C and D are denoted a, b, c and d
respectively then _BIP_ is organized like
abcdabcdabcdabcdabcdabcdabcdabcdabcd... line 1
abcdabcdabcdabcdabcdabcdabcdabcdabcd... line 2
abcdabcdabcdabcdabcdabcdabcdabcdabcd... line 3
...
abcdabcdabcdabcdabcdabcdabcdabcdabcd...
abcdabcdabcdabcdabcdabcdabcdabcdabcd...
BIP can be read with the following pseudo-code program
FOR EACH line
FOR EACH pixel
FOR EACH band
I[pixel, line, band] = get_pixel(input);
_BIL_ looks like
aaaaaaaaaaaa... band 1, line 1
bbbbbbbbbbbb... band 2
cccccccccccc... band 3
dddddddddddd... band 4
aaaaaaaaaaaa... band 1, line 2
...
BIL can be read with the following pseudo-code program
FOR EACH line
FOR EACH band
FOR EACH pixel
I[pixel, line, band] = get_pixel(input);
_BSQ_ shows
aaaaaaaaaaaa... line 1, band 1
aaaaaaaaaaaa... line 2
aaaaaaaaaaaa... line 3
...
bbbbbbbbbbbb... line 1, band 2
bbbbbbbbbbbb... line 2
bbbbbbbbbbbb... line 3
...
cccccccccccc... line 1, band 3
cccccccccccc... line 2
cccccccccccc... line 3
...
dddddddddddd... line 1, band 4
dddddddddddd... line 2
dddddddddddd... line 3
...
BSQ can be read with the following pseudo-code program
FOR EACH band
FOR EACH line
FOR EACH pixel
I[pixel, line, band] = get_pixel(input);
Of course others are possible, like the old _EROS BIP2_ format (for
four band MSS images) where the image is first divided into four
strips. EROS BIP2 strips
Then each strip is stored like
aabbccddaabbccddaabbccddaabbccdd... line 1
aabbccddaabbccddaabbccddaabbccdd... line 2
...
To decode one strip the following pseudo-code can be used
/* The '%' character is the modulo operator */
/* Note that operations on 'i' are integer operations! */
/* Copyright 1994 by W.H. Bakker - ITC */
FOR EACH line
FOR i=0 TO BANDS*WIDTH
I[(i/8)*2+i%2, line, (i/2)%4] = get_pixel(input);
Subsequently, the strips must be glued back together.
_________________________________________________________________
------------------------------
Subject: Basic Processing Levels
What are the different types of image I can download/buy?
_Very brief - needs a proper entry_
Raw data (typically Level 0)
(as with other levels, annotated with appropriate metadata).
Only useful if you're studying the RS system itself, or data
processing systems
Processed Images (typically Level 1, 2)
Processing includes:
+ Radiometric correction - compensating for known
characterisitcs of the sensor.
+ Atmospheric correction - compensating for the distortion
(lens effect) of the atmosphere.
+ Geometric correction - referencing the image to Lat/Long on
the Earth's surface, based on the satellite's position and
viewing angle at the time of the acquisition. Uses either a
spheriod model of Earth or a detailed terrain model; the
latter enables higher precision in hills/mountains. Requires
Ground Control Points (GCPS: points in the image which can be
accurately located on Earth) for high precision.
The various part-processed levels are suitable for a image
processing studies. Most Remote Sensing and GIS applications
will benefit from the highest level of processing available,
including geocoding.
Geocoded Projected Imagery (typically Level 3)
The image is mapped to a projection of the Earth, and in some
cases also composited (ie several images are mosaiced to show a
larger scene).
Browse Images
Images you can download from the net are likely to be browse
images. These are typically GIF or JPEG format, although a
number of others exist. Whilst providing a good idea of what is
in an image, they are not useful for serious applications. They
have the advantage of being a manageable size - typically of
the order of 100Kb-1Mb (compared to 100Mb for a full scene) and
are often available free. A browse version of any image (except
raw data) can be made.
Stereopairs
Multitemporal Images
------------------------------
Subject: Is there a non-proprietary format for geographical/RS images?
Is there a non-proprietary format for geographical/RS images?
The GeoTIFF format adds geographic metadata to the standard TIFF
format. Geographic data is embedded as tags within an image file.
For a detailed description, see the spec. at
http://www-mipl.jpl.nasa.gov/cartlab/geotiff/geotiff.html
------------------------------
Subject: Do I need geocoded imagery?
Do I need geocoded imagery?
In a recent discussion of mountain areas, John Berry
(ej10jlbs@shell.com) wrote:
The problem that Frank has is that he is working in an area without
adequate maps: therefore, he cannot geocode his Landsat using a DTM, because
the data available is neither detailed enough or accurate enough to use as an
input.
He can georegister the imagery using using one or two accurately
located ground control points and the corner-point positions given in the
image header: these are calculated from ephemeris data of, usually, unknown
accuracy (within +/- 1 km), but internal image geometry is good so an x,y
shift and a (usually) very small rotation can take care of everything to
better than the accuracy of his maps. Positions used should be
topographically low, and at the same elevation. GPS is the best solution, as
someone else pointed out, if Frank can get in the field.
The next problem is the parallax error introduced by the high relief.
In his situation, the only answer* is to get SPOT stereopairs and make a DTM or
DEM from them. Except in the case of very narrow gorges or slopes steeper
than 60 deg. there should be few problems with carefully chosen images (high
sun angles, etc). ERDAS has an excellent module for doing this. However, I
doubt that Frank has the budget. I believe ERDAS`s Ortho module would then
allow Frank to make an Ortho image that would be a perfectly good map.
*there may be some LFC or Russian stereo coverage in this area, which
would be a lot cheaper than SPOT but would require the use of analog stereo
comparators (probably).
Even if there were good topographic contour maps for all of Frank's
area, the cost of digitising these and turning them into a usable DTM would
probably be prohibitive (though there are outfits in Russia who might be able
to quote a price affordable to a large western company).
------------------------------
Subject: Imaging Instruments
Imaging Instruments
How do Remote Sensing Instruments work?
If you put a camera into orbit and point it at the Earth, you will get
images. If it is a digital camera, you will get digital images.
Of course, this simplistic view is not the whole story.
Digital images comprise two-dimensional arrays of pixels. Each pixel
is a sensor's measurement of the albedo (brightness) of some point or
small area of the Earth's surface (or atmosphere, in the case of
clouds). Hence a two-dimensional array of sensors will yield a
two-dimensional image. However, this design philosophy presents
practical problems: a useful image size of 1000x1000 pixels requires
an array of one million sensors, along with the corresponding
circuitry and power supply, in an environment far from repair and
maintenence!
Such devices (charge coupled deices) do exist, and are essentially
similar to analogue film cameras. However, the more usual approach for
Earth Observation is the use of tracking instruments:
Tracking Instruments
1. A tracking instrument may use a one-dimensional array of sensors -
one thousand rather than one million - perpendicular to the
direction of the satellite's motion. Such instruments, commonly
known as pushbroom sensors, instantaneously view a line. A
two-dimensional image is generated by the satellite's movement, as
each line is offset from its predecessor. If the sampling
frequency is equal to the satellite's velocity divided by the
sensor's field of view, lines scanned will be contiguous and
non-overlapping (although this is of course not an essential
property).
_btw, would the above be better expressed in some ASCII
representation of mathematical notation?_
2. Another approach is to use just a single sensor. It is now not
sufficient to use the satellite's motion to generate an image:
cross-track scanning must also be synthesised. This is
accomplished by means of a rotating mirror, imaging a line
perpendicular to the satellite motion. These are known as scanning
instruments. This is somewhat analagous to the synthesis of
television pictures by CRT, although the rotating mirror is a
mechanical (as opposed to electromagnetic) device.
As the sensor now requires a large number of samples per line, the
sampling frequency necessary for unbroken coverage is
proportionally increased, to the extent that it becomes a design
constraint. A typical Earth Observation satellite moves at about
6.5 Km/sec, so a 100m footprint requires 65 lines per second, and
higher resolution imagery proportionally more. This in turn
implies a sampling rate of 65,000 per second for a 1000-pixel
swath. This may be alleviated by scanning several lines
simultaneously.
Either design of scanning instrument may have colour vision (ie be
sensitive to more wavelength of light) by using multiple sensors
in parallel, each responding to one of the wavelengths required.
List of Imaging Spectrometers
http://www.geo.unizh.ch/~schaep/research/apex/is_list.html
------------------------------
Subject: What is a Sounding Instrument?
What is a Sounding Instrument?
_Answer posted by Wayne Boncyk (boncyk@edcsgw4.cr.usgs.gov) to
IMAGRS-L_
Satellite-borne remote sensing instruments may be used for more than
imaging; it is possible to derive information about the constituents
of the local atmosphere above a ground target, for example. One common
area of study is to observe atmospheric emissions in the spectral
neighborhood of the 183GHz water absorption line (millimeter-wave;
in-between microwave and thermal IR). These channels can be monitored
by an appropriate collection of narrow passband radiometers, and the
data that are returned can be analyzed to deduce the amount of water
vapor present at different levels (altitude layers) in the atmosphere.
The reference to "sounding" is an application of an old nautical term,
the investigation of the state of a medium at different depths
(original application: the ocean - specifically determination of the
depth of the ocean floor).
------------------------------
Subject: Orbits
Orbits
_Need a general entry here!_
Where can I learn about satellite orbits?
Wim Bakker has compiled a list of online references at
http://www.itc.nl/~bakker/orbit.html.
Wim adds the question _"When can *I* see a specific satellite"_, and
suggests the following pointers from his list:
* Visual Satellite Observer's Home Page:
http://www.rzg.mpg.de/~bdp/vsohp/satintro.html
* Satellite Observing Resources:
http://www-leland.stanford.edu/~iburrell/sat/sattrack.html
Satellite Orbital Elements
_Thanks to Peter Bolton (pbolton@clyde.pc.my) for this one!_
Jonathan's Space Report is at
http://hea-www.harvard.edu/QEDT/jcm/jsr.html. The introduction:
The Space Report ("JSR") is issued about once a week. It describes all
space launches, including both piloted missions and automated
satellites. Back issues are available by FTP from sao-ftp.harvard.edu
in directory pub/jcm/space/news. To receive the JSR each week by
direct email, send a message to the editor, Jonathan McDowell, at
jcm@urania.harvard.edu. Feel free to reproduce the JSR as long as
you're not doing it for profit. If you are doing so regularly, please
inform Jonathan by email. Comments, suggestions, and corrections are
encouraged.
How do I convert Landsat Path/Row to Lat/Long?
In response to this question, Wim Bakker wrote:
The SATCOV program is available by anonymous FTP from sun_01.itc.nl
(192.87.16.8). Here's how to get it:
$ ftp 192.87.16.8
Name: ftp
Password: your-email-address
ftp> bin
ftp> idle 7200
ftp> prompt
ftp> cd /pub/satcov
ftp> mget *
ftp> bye
$
If you can't use FTP, drop me a line and I will send a uuencoded version
by email.
Those of you who prefer a WWW interface can obtain it from the following URL:
http://www.itc.nl/~bakker/satcov
Don't forget to set the "Load to local disk" option.
SATCOV is a PC program for converting Path/Row numbers of Landsat and
K/J of SPOT to Lat/Lon and vice versa. Furthermore it can predict the orbits
of the NOAA satellites, although I wouldn't recommend it for this purpose!
But that's an other can of worms....
------------------------------
Subject: Ground Stations
How is satellite data recieved on the ground?
_Intro to Ground Recieving Stations contributed by Peter Bolton
_
1. GROUND RECEIVING STATIONS
This document is an introduction to Ground Receiving Station (GRS)
acquisition and processing of remote sensing satellites data such as
SPOT, LANDSAT TM and ERS-1 SAR. Ground receiving stations regularly
receive data from various satellites so as to provide data over a
selected areas (a footprints approximately covers a radius of 2500 km
at an antennae elevation angle of 5 degrees.) on medium such as
computer tape, diskette or film, and/or at a specific scale on
photographic paper. GRS are normally operated on a commercial basis of
standard agreements between the satellite operators and the
Governments of the countries in which they are situated. Subject to
the operating agreements, local GRSs sell products adapted to end
users needs, and provide remote sensing training, cartography, and
thematic applications.
2. GROUND RECEIVING STATION ARCHITECTURE
A Ground Receiving Station consists of a Data Acquisition System
(DAS), a Data Processing (DPS) and a Data Archive Center (DAC).
2.1. DATA ACQUISITION SYSTEM
DAS provides a complete capability to track and receive data from the
remote sensing satellite using an X/S-band receiving and autotracking
system on a 10 to 13meter antenna in cassegranian configuration. DAS
normally store fully demodulated image data and auxiliary data on High
Density Digital Tapes (HDDTs). However, in one small UNIX based
system, data storage can be stored directly on disk and/or
electronically transmitted to distant archives.
2.2. DATA PROCESSING SYSTEM
DPS keeps an inventory of each satellite pass, with quality assessment
and catalog archival, and by reading the raw data from HDDTs,
radiometrically and geometrically corrects the satellite image data.
2.3.DATA ARCHIVE CENTRE
The Data Archive closely related to DPS offers a catalog interrogation
system and image processing capabilities through an Image Processing
System (IPS).
3. GROUND RECEIVING STATION PRODUCTS
The GRS products can either be standard or value added products. Both
are delivered on Computer Compatible Tapes (CCTs), CD ROM, cartridges,
photographic films or photographic paper prints at scales of 1:250
000, 1:100 000, 1:50 000 and 1:25000.
i. Standard products
- SPOT-1 and 2/HRV : data of CNES levels 0, 1A, 1B, 2A
- Landsat TM : data of LTWG levels 0, 5,
- ERS-1 SAR : Fast Delivery and Complex products.
ii. Value added products
- For SPOT
. P + XS : Panchromatic plus multi-spectral,
. SAT : a scene shifted along the track,
. RE : a product made of 2 consecutively acquired scenes,
. Bi-HRV : Digital mosaic produced by assembling 2 sets
of
2 scenes acquired in the twin-HRV configuration.
. Stereoscopy : Digital terrain model (DTM) generation,
. Levels 2B, S and level 3B using DTMs.
- For Landsat TM: levels 6, S and 7.
- For ERS-1 SAR : geocoded data.
- For any instrument:
. Image enhancement and thematic assistance,
. Geocoded products on an area of interest defined by the
customer (projection, scale, geocoding and mosaicking
according to the local map grid).
4. GROUND RECEIVING STATION OPERATION
Persons needing images for thematic applications in the field of
cartography, geology, oceanography or intelligence, etc, will refer to
the station catalog in order to find out if the data are available
over the area concerned.
There are two possibilities :
The data exists.
The customer fills in a purchase order and is then provided
with the product on a medium such as CCT, film or paper print.
If the data are available in the GRS catalog, a list of the
related scenes and their hardcopies (named "quick looks") are
provided.
The data does not exist.
a) For SPOT, the customer fills in a programming request form
which is sent by GRS to the Mission Control Centre (MCC) in
Toulouse, France. MCC returns a Programming Proposal to be
submitted for approval. Upon approval, the confirmation is
returned to MCC which in turn sends a programming order to the
satellite for emitting the data during its pass over the GRS
antenna.
At the same time, MCC sends to GRS, the satellite ephemerides
for antenna pointing and satellite tracking.
In the case of SPOT, if the data does not exist within the
Station catalog but are listed in the SPOT IMAGE worldwide
catalog, GRS may request the level O product from SPOT IMAGE in
TOULOUSE in order to process it locally.
b) For other sensors, LANDSAT TM or ERS-1, the satellite
ephemerides are known at GRS and the antenna is pointed
accordingly in order to track all selected passes.
Within the GRS, the raw satellite data are received by the Data
Acquisition System (DAS), and recorded on High Density Digital Tapes
(HDDTs). HDDTs are then sent to the Data Processing System (DPS),
where an update of the Station catalog is made as well as a quick look
processing.
DPS is also in charge of automatic processing of selected raw data in
order to produce images of standard level.
Value added products with cartographic precision are produced within
DPS using interpretation workstations which must be part of an
operational Geographic Information System (GIS) combined to an Image
Processing System (IPS).
Once processed, the data, on CCT, are sent to the Data Archive Center
(DAC) where they are delivered to the customers after a quality
checking. At DAC, further processing may be applied to the data such
as image stretching, statistical analysis, DTM, or a conversion from
tape to film and paper prints in the photographic laboratory;
"customized services" may also be offered.
_________________________________________________________________
Image Interpretation
------------------------------
Subject: How can I assess my results?
How can I assess my results?
_(for basics, see Russell Congalton's review paper In Remote Sens.
Environ. 37:35-46 (1991). Think we should have a basics entry here
too!)_ Michael Joy (mjoy@geog.ubc.ca) posted a question about
Contingency table statistics and coefficients, and subsequently
summarised replies:
Second, a summary of responses to my posting about contingency table statistics
and coefficients. Basically, I need to come up with a single statistic for
an error matrix, along the lines of PCC or Kappa, but which takes into
account the fact that some miscalssifications are better or worse than others.
Tom Kompare suggested readings on errors of omission or commission.
Chris Hermenson suggested Spearman's rank correlation.
Nick Kew suggested information-theoretic measures.
Others expressed interest in the results; I'll keep them posted in future.
The responses are summarized below.
===============================================================================
Michael:
Your thinking is halfway there. Check out how to use an error matrix to get
+ errors
of Omission and Commission.
Good texts that explain it are:
Introduction to Remote Sensing, James Campbell, 1987, Gulliford Press
start reading on page 342
Introductory Digital Image Processing, John Jensen, 1986, Prentice-Hall
start reading on page 228 or so.
These are the books where I learned how to use them. Sorry if you don't have
+ access
to them, I don't know how Canadian libraries are.
Tom Kompare
GIS/RS Specialist
Illinois Natural History Survey
Champaign, Illinois, USA
email: kompare@sundance.igis.uiuc.edu
WWW: http://www.inhs.uiuc.edu:70/
============================================================================
Excerpt from my response to Tom Kompare (any comments welcome...)
These are useful readings describing error matrices and various measures we can
get from them, eg PCC, Kappa, omission/commission errors. But from these
+ readings
I do not see a single statistic I can use to summarize the
whole matrix, which takes into account the idea that some misclassifications
are worse than others (at least for me). For example, if I have two error
matrices with the same PCC, but with tendencies to confuse different categories
,
I'd like to get a ststistic which selects the 'best' matrix (ie the best image)
.
One simple way I can think of to do this is to supply a matrix which gives
a 'score' for each classification or misclassification, and then multiply each
number in the error matrix by the corresponding number in the 'score' matrix.
So a very simple example of such a matrix might look like this:
Deciduous Conifer Water
Decid 1.0 0.5 0.0
Conifer 0.5 1.0 0.0
Water 0.0 0.0 1.0
In this notation, the 'score' matrix for a PCC statistic would be a diagonal
matrix of "1". Obviously there are a number of issues for me to think about
in using such a matrix, eg can you 'normalize' the score matrix? Can you
use it to compare different matrices with different numbers of categories?
An obvious extension to this would be to apply this idea to the Kappa
statistic as well.
===========================================================================
Hi Michael;
Spearman's rank correlation is often used to test correlation in a situation
where you are scoring multiple test results. You might be able to adapt
it to your problem.
Chris Hermansen Timberline Forest Inventory Consultants
Voice: 1 604 733 0731 302 - 958 West 8th Avenue
FAX: 1 604 733 0634 Vancouver B.C. CANADA
clh@tfic.bc.ca V5Z 1E5
C'est ma facon de parler.
=========================================================================
Hi,
Your question touches on precisely the field of research I'd like to be
pursuing, if only someone would fund it:)
> Hi,
> I'm comparing different datasets using contingency tables, and I would
> like to come up with summary statistics for each comparison. I am using
> the standard PCC and Kappa, but I'd also like to come up with a measure
> which somehow takes into account different 'degrees' of misclassification.
> For example, a deciduous stand misclassified as a mixed stand is not as
> bad as a deciduous stand misclassified as water.
I would strongly suggest you consider using information-theoretic measures.
The basic premise is to measure information (or entropy) in a confusion matrix.
I can send you a paper describing in some detail how I did this in the
not-totally-unrelated field of speech recognition.
This does not directly address the problem of 'degrees of misclassification' -
just how well it can be used to do so is one of the questions wanting further
research. However, there are several good reasons to use it:
1) It does address the problem to the extent that it reflects the statistical
distribution of misclassifications. Hence in two classifications with
the same percent correct, one in which all misclassifications are between
deciduous and mixed stands will score better than one in which
misclassifications are broadly distributed between all classes.
Relative Information is probably the best general purpose measure here.
2) By extension of (1), it will support detailed analysis of hierarchical
classification schemes. This may be less relevant to you than it was
to me, but consider two classifiers:
A: Your classifier - which for the sake of argument I'll assume has
deciduous, coniferous and mixed woodland classes.
B: A coarser version of A, having just a single woodland class.
Now using %correct, you will get a higher score for B than for A - the
comparison is meaningless. By contrast, using information (Absolute,
not Relative in this case), A will score higher than B. You can
directly measure the information in the refinement from B to A.
> In effect I guess I'm
> thinking that each type of misclassification would get a different 'score',
> maybe ranging from 0 (really bad misclassification) to 1 (correct
> classification).
I've thought a little about this, as have many others. The main problem is,
you're going to end up with a lot of arbitrary numerical coefficients, and no
objective way to determine whether they are 'sensible'. Fuzzy measures can
be used, but these are not easy to work with, and have (AFAIK) produced
little in the way of results in statistical classification problems.
> I can invent my own 'statistic' to measure this, but if there are any such
> measures available I'd like to use them. Any ideas?
Take the above or leave it, but let me know what you end up doing!
Nick Kew
nick@mail.esrin.esa.it
============================================================================
--
Michael Joy mjoy@geog.ubc.ca
University of British Columbia, Vancouver, B.C., Canada
------------------------------
Subject: Is there a program to compute Assessment measures, including Kappa coe
fficients?
Is there a program to compute Assessment measures, including Kappa
coefficients?
Nick Kew's assess.c (ANSI C source code to compute several assessment
measures, including PCC, Kappa, entropy and Mutual and Relative
Information) is available for download from the WebThing site,
http://pobox.com/%7Esatfaq/ or from the satfaq autoresponder (mail to
satfaq@pobox.com with subject line "send assess.c").
_Old reference to Dipak Ram Paudyal's kappa program deleted, as the
FTP server is apparently no longer available._
------------------------------
Subject: How good are classification results in practice?
How good are classification results in practice?
The following detailed commentary was posted by Chris Hermansen
(clh@tfic.bc.ca).
Mike Joy posted a question regarding irregularities between two
classifications, one derived from manual interpretation of
large-scale aerial photography, the other from a supervised and
enhanced spectral classification of Landsat TM imagery.
I've read several of the responses, and I just thought it time
to kick in my $0.02 worth, since I am quite familiar with both
of the classifications with which Mike is working.
First, Peter Bolton rattles off his experience in tropical forests
and chastises Mike for discovering what should have been obvious.
Well, Peter, the boreal forest is a much different beast than
what you're used to in Malaysia (I can attest from firsthand
experience in both cases). Classification from remotely sensed
data is generally quite reliable in the boreal forest, especially
given the vegetative nature of the TM-derived classification
that is Mike's second dataset. Detecting predominantly deciduous
from predominantly coniferous stands is (spectrally speaking)
pretty straightforward. Problems arise in mixedwood stands,
however, since the nature of the classification of proportion
is not necessarily the same and in any case any aggregative
techniques applied to the TM image prior to classification (eg
smoothing) could significantly alter the proportional balance.
Also, depending on the proportion of deciduous in a predominantly
coniferous stand, and the spatial distribution of deciduous trees
within that stand, the classifier may have difficulty detecting
the differences between mixedwood and younger pure coniferous
types. Furthermore, deciduous stands with coniferous understory
are classified as deciduous in Mike's first dataset but may
easily be interpreted as mixedwood stands in the TM image.
Secondly, on the subject of incorporation of field data, Mike's
second dataset has some ground truthing incorporated in the
classification.
Thirdly, on the subject of large numbers of classes in some
people's TM-derived classifications, remember that in many cases
these additional classes are derived by incorporating other
datasets (field measurements, other digital map data, DEM
information, etc). The people I've seen most test this envelope
are the folks at Pacific Meridan Resources; their TM-derived
datasets form only the first step of several. As Vincent
Simonneaux points out, most people stop at the first step.
So, in response to Mike's original questions:
> 1) Is it reasonable to expect a TM-based classification to accurately
> distinguish Coniferous and Deciduous forest? The area I am dealing
> with is boreal mixedwood forest in northeren Alberta, Canada. I had
> expected that the classification should at least be able to do this.
On the face of it, yes. But! You must ensure that your definition of
Coniferous and Deciduous forest is exactly the same in both cases (and
the prevailing definitions in use in Alberta don't exactly help out in
this case).
> 2) Do people out there have similar experiences, i.e. the actual
>classification
> accuracy being very much lower than the reported results, or major
> differences when comparing with different source of information?
Of course, this is a possibility; the most unreliable classes may
interfere in a nasty way between to datasets. You really need to ensure
that you are sampling the same population in both cases; then you need
to examine the distribution of errors among classes in both cases. In
your first dataset, you don't really have error estimates with which to
work.
> I
> understand that an air-photo-based forest inventory and a TM satellite
>image
> are measuring different things, and that I shouldnt expect perfect
>agreement,
> but I would have thought they could agree roughly on the overall area of
> Coniferous or Deciduous forest. Ditto for two similar TM-based
> + classifications.
Once more, not necessarily. See the points above on coniferous understory
in deciduous stands and the basic definitions of coniferous/deciduous
split.
There are, of course, really obvious errors that can occur, like using
pre-leaf or post-leaf images when trying to locate deciduous stands...
Sorry to go on at such length about this; I hope that my comments are of
interest to some of you.
------------------------------
Subject: I need to classify a mosaic of several images. How best to do it?
I need to classify a mosaic of several images. How best to do it?
David Schaub (dschaub@dconcepts.com) posted a question on this. Here
is his summary of replies:
Dear Netters,
Some time ago I posed a question to this list with regards to classification,
rectification, and mosaicking. My original question was as follows--
>Hello,
>We need to georectify, mosaic, and classify several (3 or 4) Landsat TM
>scenes using ERDAS Imagine. The classification will need to show major
>land cover categories, such as bare ground, grassland, shrubby range,
>built-up, coniferous forest, broad-leaf forest, water, etc. In the past
>when we have done this the seams between images are quite evident in the
>classification. We would like to minimize differences between images, yet
>be asaccurate as possible in the classification of each image.
>My main questions are these -- Should we classify each image separately
>and then mosaic them, or should we mosaic the images first and then
>classify them? Can georectifying the images effect the classification?
>You can assume that images along a path will have the same acquisition date,
>however scenes on adjacent paths will have different dates (at least by two
>weeks). I will post a summary. Thanks in advance for your opinions :-)
This quickly generated a flood of responses. While there wasn't complete
agreement, the majority of respondents believed that I should first classify
the images, then do the rectification and mosaicking. Nearest neighbor
should be used when rectifying the classified image (or if the image data
are rectified before classification). Thanks to all who responded!! Comments
are summarized below:
David Schaub
dschaub@dconcepts.com
*******************************************************************
I have done the same things you are attempting to do for my thesis work.
I think the best course of action would be to classify the images first, then
rectify the images and then merge or mosaic the images. Rectifying the images
before you classify may distort the spectral characteristics of pixels and
thereby influence your classification. Furthermore, the smaller the area you
are classifying, the more accurate the classification will be, so if you
mosaic a large area and then attempt to classify the mosaiced image, there
will be more confusion possible based on the heterogeneity of a larger area. I
hope this helps, contact me if i can be of further assistance.
David Smith
*************************************************************
Here's my 2c for what it's worth...
I classify TM scenes separately and then mosaic the classifications. My
classifications almost never have a seam in them...If there is a seam
it is usually due to the difference in the date of the scene. You have
to be careful though... you need to use the same method of classification
(plotting out feature spaces and elipses helps) for overlapping scenes.
Sometimes this is why people use the other method...
If you're going to do this the other way round...i.e. mosaic and then classify
scenes you will have to calibrate the scenes to radiance and then use some
kind of atmospheric correction before mosaicking them. This should in theory
minimize the difference in the spectral information between scenes....I would
avoid using any kind of histogram equalization ...although it may look nice,
you are loosing the original pixel information.
\\. _\\\_____
\\\ /ccccccc x\ Fiona Renton, GIS and remote sensing analyst
>>Xccccccc( < CALMIT, Conservation and Survey Division
/// \ccccccc\_/ University of Nebraska-Lincoln
'' ~~~~ renton@fish.unl.edu
**********************************************************************
What sort of classification? Pixels? Clusters? Polygons?
Higher-level features? If your classification units are homogenous
and shape is not important, you should clearly do it before mosaicing.
If not, you have a genuinely interesting problem, and will probably have
to your own research (starting at your local academic library, assuming
there is one :-)
Nick.
*********************************************************************
Geo-rectification will have a small effect on classification due to the
resampling process. I can't help to much on classification part,
because that is not my area, but my feeling is that mosaicking
non-classified images may be easier than trying to match features in a
classified image.
Ok, this is my area. You can not assume that images on the same path
are imaged on the same day, However, they could be. You should be able
to check the meta-data to find out if they were. The next path west
could have been imaged 7 days after the path of interest or 9 days
before and the next path east could have been imaged 9 days after the
path of interest or 7 days before, again check the meta-data. The next
chances are to add 16 days on to those numbers (i.e. 7 + 16).
This is true for Landsat 4 and 5 only (will be true for Landsat 7).
Chuck
wivell@edcsnw38.cr.usgs.gov
************************************************************
Yes the georectification process will affect the classification
results. My suggestion is to classify each individual image first and
then mosaic them together. I have done this before and it works well.
If you mosaic first and then classify you have to calibrate the data,
apply radiometric corrections etc... Not worth the trouble in my
opinion, and you probably won't get any good results.
The resampling technique (convolution) will affect the radiometric value
of the image and may not be suitable for adequate identification
aftrewards. To avoid visible seams, just go around the areas, try to
contour the natural groupings (classes after classification)
To resume, in my opinion, if you want good accurate results: Classify first
and after mosaic.
Francois Beaulieu
************************************************************
You definitely want to mosiac the 4 images first (into one file) and
then run the classification on that. Because of subtle differences in
the radiometric characteristics of each image, the classes in separate
classifications will rarely "line up" perfectly when mosaicked afterward.
> Can georectifying the images effect the classification?
Yes it can, depending on the resampling technique you use. When
rectifying the images, use Nearest Neighbor resampling as that
will ensure that original pixel values are used to create the
new rectified dataset. (Bilinear or Cubic will average the
original data, resulting in slight degradation.) I would:
1) Rectify the four images (use Nearest Neighbor)
2) Contrast balance them, using for example Histogram
Matching or another technique.
3) Mosaic the four contrast balanced scenes into one file.
4) Run the classification.
I hope this helps.
Eric Augenstein
Manager of Training Services
*************************************************************
In general you can't depend on the DN values from one image to the next
to be related. You should classify before your mosaic - in other words
mosaic the classification, not the images. Otherwise you mix unrelated
DN values into a signal classification which would be wrong.
Classification may be affected by geo-rectification. If the
geo-rectified image has the same pixels and pixel values as the
original, the classification should not be affected. However, this is
an unreal assumption. A geo-rectified image will almost always have
resampling - which means that pixels are either dropped or replicated -
unless a filter is applied (like bilinear or cubic convolution) in
which case the pixel values change as well.
If the classifier is single pixel based (like isodata) then the
classification is only affected by the resampling as the sigatures are
affected by the replication or dropping of values. If the classifier is
regional or global (like multi-resolution/multi-scale classifiers, or
region linking) then the classifiers may be affected to a greater
degree.
You can classify before or after geo-rectifiction and the results
will not be vastly different.
But the bottom line to mosaic at the very end.
Michael Shapiro mshapiro@ncsa.uiuc.edu
NCSA (217) 244-6642
605 E Springfield Ave. RM 152CAB fax: (217) 333-5973
Champaign, IL 61820
********************************************************
Re Michael Shapiro's posting,
There is no doubt that that you cannot depend on the DN values from one
image to the next (especially with images from adjacent paths which are
taken on different dates (see Chuck Wivell's posting).
However mosaicing images which have been classified seperately may produce
unusual results ie trying to match classes from different images.
A suggestion would be to first try some kind of atmospheric correction on
the images, mosaic them and then classify them together. Assuming
i) you can do a credible atmospheric condition (using Dark Pixel
Substraction, Band Regression etc) plus, perhaps, correct the images to a
constant solar elevation angle
ii) the images from different paths were not taken on widely different
dates and
iii) (linked to ii) the ground conditions are similar for the images
from different paths
then the DN values between images should be comparable.
Euan
************************************************************
We are currently doing a statewide land cover classification for Mississippi
using TM scenes (10 of them). My responses for your questions:
1. We classified each scene separately - mainly because the dates differed
and in the cases where we had adjoining scenes taken on the same day, it
was decided that classifying a full scene was a big enough task in both
computer and human resources. If you had subscenes, it would not be too
bad. I would advise against mosaicking scenes before classifying - your
signatures for the same landcover class in the other scene(s) would be
different and it would be a nightmare. Matching techniques that changed
image pixel values would change your original data and corrupt your
classification.
2. We also georeferenced each scene before classification for the following
reasons:
- georeferenced ancillary data sources (roads, streams, NWI, etc) were
used - including leaf-off TM scenes already in-house.
- the need to have maps to take into the field for pre and post
classification checks.
We used nearest neighbor. This doesn't change pixel values but just
moves them to a different location. In our case the image statistics
were unchanged after georectification although it is probable that some
pixels may be dropped or replicated (but when you georeference the
classified image, those same pixels are going to be affected anyway).
Bottom line would be to classify each scene separately. I would georeference
each TM scene first - when the classiciations are completed, stitching is
easy.
Jim
************************************************************
Our lab has had luck using regression techniques to mosaic the three
bands together. Using ERDAS imagine, the steps are:
1) create an image where the two scenes overlap (this is best
done with modeller, not layerstack: layerstack only
uses the geographical boundaries, whereas you want to
have the area where there are values in both images
2) Use the Accuracy Assessment module to create random points
on the image and remove those points which lie in cloud
or shadow.
3) Export the X,Y coordinates from the random points and use
these as a point file in the Pixel-to-Table function.
Use the overlap image as the output image (make sure
you have all the bands you want to regress (ie. image
one's band 3,4,5 on top of image 2's 3,4,5
4) You now have a set of points that can be imported into
any standard statistical package. You need to have the
values from the "larger" or primary image be the Y values
and the other image be the X value (I'm told the correct
statistical term is that the Y is the master and the X
is the slave).
This should create a seamless image. Obviously, the closer the B
number in the Y= bx + constant equation is to 1, the less you are
transforming the values of your slave image. We have also tried doing
classifications of each image first, but the results have been
disappointing.
Regards,
Sean Murphy
University of Maine
********************************************************************
