geoinformatics 2006 vol02
TRANSCRIPT
-
7/27/2019 geoinformatics 2006 vol02
1/51
Educate Industry to Dangers InhereER Mapper Can Leave Patent Case Behind
Educate Industry to Dangers Inhere
Finally, after a period of six years, the
dispute between ER Mapper and
Lizardtech has been settled.
GeoInformatics recently had an interview
on this topic with Stuart Nixon, founder of
Earth Resource Mapping (ER Mapper).
By Sonja de Bruijn
The patent case (U.S. Patent No.
5,710,835, related to image compression
technologies) dates back to 99. Therewere also other claims of LizardTech and
a lawsuit against you personally.
Looking back on this now: how did this
influence you and the company you
founded?
Beginning in 1999, LizardTech launched three
essentially identical US Federal Court legal
actions against us; two against the company
and one against myself. LizardTech made a
whole range of claims. Their core allegation
was that we infringed the '835 wavelet com-pression patent that LizardTech had licensed
from the US Federal Government Los Alamos
National Laboratory. The court threw out their
legal action against me.
The two actions against the company were
handled as one by the courts and the final
aspects were wrapped up recently when
LizardTech's appeal and en banc appeals were
rejected. Significantly, we won the cases on
Summary Judgement. In other words:
LizardTech's three legal actions never went to
trial. We were able to show the Federal Judge
presiding over the case that LizardTech had no
basis for their patent infringement claims.
This was my first direct experience with US
patent law. It was an eye opener to see how a
litigious company can use the US legal system
as an anti-competitive tactic. We were heart-ened by the tremendous and positive encour-
agement people in the industry gave us during
the nearly 6 years it took us to defend our-
selves and the ISO JPEG 2000 standard from
March 20066
LizardTech's actions. Having said that, I really
don't think that most people today outside the
US appreciate just how all pervasive and dam-
aging the US patent system is. It does not just
affect companies in the US. Because the US
patent system enables companies to claim a
total percentage of product revenue that might
arguably use a patent (rather than a percent-
age measured by the effective value of that
patent to the products), companies have
strong incentive to launch speculative patent
actions. Probably the only thing preventing
total mayhem is that the major technology
companies all have very large patent portfolios.
Any attempt by one major player to start large
scale patent litigation against another large
player would turn the entire US technology
industry into instant grid-lock. However, as wehave seen in this instance, actions by smaller
companies can still cause tremendous damage
to users, competitors, and industry.
Please indicate some of the points that
made the Court decide in favour of ER
Mapper.
There were two key points. First, the Court
found that our way of performing the Discrete
Wavelet Transformation (DWT) is quite differ-
ent and so did not infringe. Second, the Courtfound that Claim 21 of the '835 patent
licensed by LizardTech is invalid because it
did not describe seamless DWT tiling and so
offered nothing new over prior art.
Interv iew
Stuart Nixon, founder of ER Mapper.
Comment by LizardTech:
After careful consideration we have decided not to pursue this case any further.
We realize this case has gone on for a long time and we must focus on what is
most important to us at this time and that is getting back to competing in the mar-
ketplace.
LizardTechs commitment in the geospatial community is to its customers and part-
ners by bringing innovative and groundbreaking geospatial software to meet the
needs of GIS professionals by helping them overcome the challenges associated
with massive satellite imagery and aerial photography.
Carlos Domingo, President and CEO LizardTech.
-
7/27/2019 geoinformatics 2006 vol02
2/51
Is the case really finished now or can
the petition by Lizardtech still be of
influence? What is the current status?
We believe the case is now really finished
and totally over with. After we won some
time ago on Summary Judgement,
LizardTech appealed to the US Federal
Circuit Court of Appeals. After their appeal
was rejected, LizardTech then petitioned
for an en banc rehearing for Claim 21 only.
The en banc hearing was also denied.
Given that the Federal District Court and
Appeals Court both rejected LizardTech'sarguments, it seems unlikely that the
Supreme Court would be receptive to any
final appeal by LizardTech (assuming they
were to try).
The US Court of Appeals has affirmed
the judgement of the United States
District Court for the Western District of
Washington that ER Mapper does not
infringe the 835 patent and that part
of the patent is invalid. What exactly
does the latest mean?
Essentially, it means that JPEG 2000, ECW
and other related wavelet formats are no
longer under threat from LizardTech's patent
Please comment on the security of theISO JPEG 2000 standard, clients and
ER Mapper as a company. Which con-
cerned you most and concerns you most
now?
Clients need long-term security for use, stor-
age and access of their valuable imagery
assets. The ISO JPEG 2000 addresses these
needs. This is why we strongly endorse and
support the standard, and the reason why we
defended it against LizardTech's actions. It is
also why we spent years developing the ECWJPEG 2000 Software Development Kit and
then released it for free. For this reason we
released full source code for the SDK. We also
try to make easy imagery access as wide
spread as possible through all software prod-
ucts in the industry, not just our own. In
short, we realize that the reason we are in
business is to support and add value to our
clients and partners.
During the litigation, our biggest concern was
the endless drain of litigation costs which
would have been put to better use in improv-
ing software for our clients. Our concern now
is to educate the industry to the dangers
inherent in US software patents, not just for
the US, but also for Europe and Asia.
How does the outcome of the case influ-
ence the future of ER Mapper and its
products?
Other than reducing uncertainty for the JPEG
2000 standard and of course for our ECW for-
mat, it does not change things. ER Mapper is
very focussed on the logical progression ofimagery use, which is imagery deployment
across the Internet by enterprises.
I still smile and shake my head in wonder
whenever I see terabyte image mosaics being
served over the Internet using our technology
and accessed by users all over the world in
all sorts of products and applications. It is
exciting and fun. I am proud to have played a
part in the development of the geospatial
imagery industry.
Sonja de Bruijn ([email protected])is editorial manager of GeoInformatics.
More information can be obtained at
www.ermapper.com and www.lizardtech.com .
litigation.Claim 21 of the '835
patent is missing
the crucial add-back
step that resulted in
generation of a
seamless DWT from
individual image
tiles. Had the claim
stood then JPEG
2000's non-seam-
less tiled DWT
method would haveinfringed. This is
perhaps why
LizardTech contin-
ued to try to keep
Claim 21 from being
invalidated, even
after giving up on
appeals against our
DWT method non-
infringement. In
their petition for
an en banc hearing by the entire US Federal
Circuit Court of Appeals, LizardTech argued
"To reach this novel result, the panel first
read a non-existent "seamless" requirement
into Claim 21 [...]" (page 4, line 1, emphasis
added).
So you can see that in LizardTech's own
words they believed Claim 21 covered
non-seamless DWT tiling. JPEG 2000 does
non-seamless DWT. I frankly find it con-
temptible that, despite LizardTech's market-
ing spin claiming they were not after JPEG
2000, their legal actions spoke otherwise.
Not only did this case have a long
history but also an expensive one. Could
you give a global indication of the total
costs?
Total damage to the industry is hard to
quantify. Certainly it cost us many millions
of US dollars to defend. It probably cost ER
Mapper tens of millions in lost business
and lost opportunities. As for the industry
itself, the litigation introduced considerable
uncertainty for a long time, and held backwide spread use of large geospatial
imagery. My own feeling is the total indus-
try cost was well over a hundred million
dollars.
March 2006Latest News? Visit www.geoinformatics.com 7
t in US Software Patentst in US Software Patents
Interv iew
-
7/27/2019 geoinformatics 2006 vol02
3/51
Reality-based 3D City ModelsCyberCity-Modeller and Database
Reality-based 3D City Models
For geomatic applications and realistic visualization the latest urban cartographic
databases contain 3D building models. Based on the application scale and requested
details, CyberCity AG creates 3D city models from aerial / satellite imagery and
laserscanner data.
By Daniela Poli
IntroductionToday the demand for the generation and
realistic visualization of 3D urban environ-
ments is growing in many geomatic appli-
cations. First there was the move from a
2D to a 2.5 D representation of the reality
with the introduction of terrain models.
Now GIS users are looking for the descrip-
tion of buildings as 3D vector data in order
to create a virtual environment which is
more and more similar to reality. The third
dimension represents a fundamental infor-
mation for efficient disaster simulations
(like earthquakes and flooding), urban and
environmental planning, and building moni-
toring. Other examples are telecommunica-
tion planning, pollution distribution analy-
sis, microclimate investigations and securityevaluation.
In the field of visualization, the added
value provided by a 3D object in compari-
son to the corresponding 2D plan is incom-
mensurable. By flying through the 3D city
models (see the interactive model of
Salzburg, Austria, (1)) the user can recog-
nize the location and get a true impression
of the presence of the buildings around
him. For tourism purposes, realistic models
with a high level of rendering are recom-
mended.
Building Geometric ComplexitiesAccording to the application and the project
scale, the 3D city models must fulfil specific
accuracy requirements in the geometry and in
the texture. Memory size is an essential issue
too. At CyberCity three different levels of geo-
metric complexity are generated for city mod-
els, see Figure 1: (A) block models, (B) main
roof structure and (C) detailed roof structures.
Of course, the more details, the higher the res-
olution of the input data.
The data source can consist of stereo aerial
images, satellite stereopairs and laserscanner
data. In practise, detailed roof structures can
be generated from aerial stereo images with a
scale 1:9,000 or smaller or from dense laser-
scanner measurements with the support of
orthophotos. Examples can be found in
Figures 2 and 3. With the highest possible res-
olution for satellite images available on the
market (Quickbirds pixel size is better than70cm) main roof structures can be modelled.
In addition overhanging roof structures (D) can
be generated by combining the information
from the sky with the planimetric building
footprints.
CyberCity ApproachUsing a reality-based semiautomatic pho-
togrammetric approach, CyberCity generates
3D city models from aerial images, laserscan-
ner data and satellite images with its dedicat-
ed software CyberCity-Modeller (CCM).
The modelling approach is the following. In
case of aerial and satellite images, see Figure
4, the relevant roof points are first measured
three-dimensionally in a photogrammetric sta-
tion following specific rules, then they are
imported as a point cloud and automatically
fitted with roof faces. For common types of
roof that follow geometric constraints like right
angles and parallel lines, intelligent measure-
ment rules were developed to reduce the
number of points that are required to be mea-
sured to create an object. Therefore this point
cloud coding reduces manual labour and
keeps the independence of the photogram-metric workstation for 3D measurement and
building generation.
The vertical building walls extrude from the
intersection of the roof polygons with the
Digital Terrain Model or alternatively through
8
Art i c le
Figure 1: Different levels of geometry complexity.
March 2006
Figure 2: 3D City Model from aerial images: Los Angeles, with detail in Little Tokyo. Automatic texturing using
aerial images Real-Time visualization with the VRGIS TerrainView (ViewTec AG).
Figure 3: Detailed roof structures from laserscanner
data: the 3D model of Bonn. Real-time visualization
with VRGIS TerrainView (ViewTec AG). Published with
the approval of the Landesvermessungsamt
Nordrhein-Westfalen.
-
7/27/2019 geoinformatics 2006 vol02
4/51
back-projection of the building footprints from
the cadastre. In the latter case, overhanging
roofs are automatically generated. Special
modules for quality control allow the improve-
ment of the geometry due to inaccuracies in
the measurements, for example right angles,
parallel lines, planar faces, and correction of
overlappings and gaps.
It is also possible to introduce geometric
attributes like area, volume and compute them
automatically for any buildings of interest.
More technical details are described in (3).
The accuracy of the 3D models is dependent
on the image scale (governed by the flight alti-
tude) and is about 0.1 - 0.2 m using a repre-
sentative fraction scale of 1:5,000.
Concerning satellite images, CCM can handle
orientation is not known, usually in case of
oblique aerial images, the images can be ori-
ented by measuring six tie points in the 3D
model and in each image.
Once the orientation is known, the software
automatically projects each faade or polygon
to the images. Among the available textures
for the selected faade, the software chooses
the best one according to the number of pix-
els and the occlusion grade through neigh-
bouring buildings. If the texture selected auto-
matically by the software is not entirely
pleasing, the user can view the available tex-
tures from the other images interactively on
the polygon in the 3D model and choose the
preferred one.
Each faade image can also be opened directly
in an image processing software for editing
and retouching to remove any obstacles or
correct the radiometry. The changes can be
seen interactively on the 3D model. Thanks to
the high degree of automation, large size 3D
city models can be mapped with realistic tex-
ture in a short time with this approach.
The last and most realistic option for faadetexturing is the manual mapping of terrestrial
photographs with high resolution and quality;
the quality and rendering achieved are very
high. The images are oriented by selecting the
face of the polygon in the 3D model and the
orientation data in Rational Polynomials
Coefficients (RPC) formats and apply correc-
tions with different functions, such as shift,
rotation, and affine.
Roof and Faade TexturingIn CCM roofs are textured automatically using
aerial images, orthophotos or satellite images.
For faades, three methods are followed,
according to the available data and the ren-
dering level to be achieved, see Figure 5. If no
data (aerial or terrestrial images) are available,
it is possible to map each faade choosing the
most suitable texture in the available library.
This method can be automated for groups of
faades but produces models which are not
reality-based.
Following an alternative approach, in CCM it is
possible to extract the faade textures from
the aerial images previously processed for the
extraction of the building structure or from
high-resolution oblique aerial images. If the
March 2006Latest News? Visit www.geoinformatics.com 9
Art i c le
from Aerial and Satellite Datafrom Aerial and Satellite Data
Figure 4: Processing chain in CyberCity-
Modeler.
Figure 5: Methods for faade texturing.
Figure 6: Details in Quickbird images (Courtesy of
Eurimage S.p.A.)
-
7/27/2019 geoinformatics 2006 vol02
5/51
corresponding corner points in the images.
Occlusions and obstacles like cars, trees and
people are retouched in a post-process to pro-
vide a superior faade texture.
Case StudyOne of the last results achieved at CyberCity is
the generation of a 3D city model from the
high resolution satellite sensor Quickbird.
Downtown Phoenix, Arizona, was modelled
from a Quickbird stereopair kindly provided by
Eurimage S.p.A., Italy (1).
The dataset included two stereo images over
their testsite in Phoenix, Arizona, together with
DTED2 DTM and 30 GCPs measured with topo-
graphic surveying. The images were acquired
on 9th April 2004 with viewing angles 29
(forward) and 27 (backward). The mean
ground resolution was 70 cm. In these images
it is possible to distinguish both skyscrapers
and residential houses distributed in a regular
network. The appearance of a building is
shown in Figure 6.
After the image orientation, the roof points in
the downtown are about 2 square kilometres.
These were measured in stereo mode and
transformed into 3D objects in CCM. The roof
texturing was added automatically using the
original Quickbird images. As the faades werenot visible in the scenes and no other data
were available (oblique aerial images, terrestri-
al images), some faades have been mapped
with texture available in the software library.
building roofs on the ground near the 3D
models;
For engineering and architecture planning:
by exporting the 3D models in DXF formats
and working in CAD software, engineers
and architects can be used to visualize
buildings and urban areas and simulate the
impact of planned buildings.
Worldwide DatabaseApart from providing services for the modelling
of urban environment of interest, CyberCity is
generating a database of 3D models of world-
wide cities and any geomatic user will be able
to buy the 3D model for a selected area of
interest.
For the generation of highly detailed and accu-
rate models, CyberCity uses aerial images with
a scale 1:8,000 or smaller and accurate aerial
triangulation parameters and digital terrain
models. The faades are textured with both
(oblique) aerial images and terrestrial images
for the features of main interest.
The CyberCity database includes or will soon
include Los Angeles, San Diego, Las Vegas,
Paris and Barcelona. Other cities will follow.
For relevant buildings, like stadia, churches,
and monuments, detailed 3D models have
been generated using terrestrial LIDAR data. A
number of landmarks located in Germany is
currently produced in collaboration with
Harman/Becker Automotive Systems.
References(1) Eurimage S.p.A.: http://www.eurimage.com
(2) Salzburg 3D:
www.viewtec.ch/techdiv/tvocx/salzburg.html
(3) Ulm, K., 2003. Reality-based 3D city mod-
els with CyberCity-Modeler (CC-Modeler)
and laserscanner data. VI Conference on
Optical 3D Measurement Techniques -
Gruen/Kahmen (Eds), Vol.2, pp. 32-39,
September 2003, Zurich, Switzerland.
Daniela Poli ([email protected]) is Senior ConsultantPhotogrammetry & GIS in the company CyberCity AG,
Zurich. For more information:
CyberCity AG Zurich - www.cybercity.tv
CyberCity LLC Los Angeles - www.cybercityllc.com
For visualization, the TerrainView software by
ViewTec (ViewTec, 2006) was used. A screen-
shot of the resulting 3D model is shown in
Figure 7.
ApplicationsThe textured 3D city models can be exportedinto several formats and managed in different
software according to the application:
In GIS environments: an interface allows
the user to export CyberCitys 3D city mod-
els as Shapefiles for ESRI ArcGIS 9.
Additionally, the data can be stored in a
Personal Geodatabase (PGDB) or managed
in a commercial database like Oracle or
Microsoft SQL Server using the ArcSDE con-
nection. The 3D data represent an addition-
al layer for further geographic analysis, dis-
aster simulations, microclimate analysis or
for homeland security;
For real-time visualization: the models can
be exported in Open Flight (FLT) with dif-
ferent levels-of-detail (LOD) for the textures
(as a percentage of the original resolution),
and the geometry. The LOD of the geome-
try composes of a symbol (LOD 1), the
block model with flat roof (LOD 2), the
main roof structures (LOD 3) and the main
roof including all superstructures like
dormers, and chimneys (LOD 4). The visu-
alization speed can be additionallyimproved by using hardware accelerated
texture formats. It is recommended to map
the terrain with true-orthophotos instead
of standard orthophotos to avoid seeing
March 2006 11
Art i c le
Figure 7: 3D City Model of Phoenix, Arizona. Generated and textured with CC-Modeler package. Real-time
visualization with VRGIS TerrainView (ViewTec AG). Published with the approval of Eurimage, Italy.
Example of 3D landmark: the Brandenburg
Gate, generated using terrestrial LiDAR
data and textured with high resolution
digital images. 2006 Harman/Becker
Automotive Systems GmbH, CyberCity AG.
-
7/27/2019 geoinformatics 2006 vol02
6/51
Europe Advances Rapidly TowaWide Implementation of Standards
Europe Advances Rapidly Towa
Digital devices of all kinds continue to proliferate and become smaller, more powerful
and less expensive. But what is most amazing about this progress is that they are
being made to communicate with each other on an unprecedented scale.
Communication, of course, means transmitting or exchanging through a common sys-
tem of symbols, signs or behaviour. The government programs described in this
paper have adopted the common systems, or standards, that define the Internet, the
World Wide Web, and the "Geospatial Web".
By Martin Klopfer and Guenther Pichler
Non-proprietary StandardsThe success of the government programs has
much to do with the fact that the underlyingstandards are non-proprietary standards
developed by open, global consensus stan-
dards organizations. The developers of these
standards represent a broad spectrum of
organizations technology providers and
technology users - who are committed to the
standards' widespread commercial implemen-
tation.
Geoprocessing standards help organizations
leverage their investments in systems and
data. Sharing and reusing data helps organi-
zations decrease costs and gain or provideaccess to more and better information. Open
standards also enable selection of the best
software tool for each job and reduce technol-
ogy and procurement risk, like the risk of
being bound to one vendor. A 2005 study
commissioned by the US National Aeronautic
and Space Administration (NASA) found thatprograms based on open standards have a
25% better return on investment than pro-
grams based largely on vendors' proprietary
interfaces and formats.
Organizations around the world are moving to
"open architectures", which means high level
information system designs based on open
standards, and "service oriented architec-
tures". The latest are high level information
system designs which involve computing pro-
cesses communicating with computing pro-
cesses on other servers, usually throughstandards-based interfaces and encodings.
The initiatives described in this article show
that this transition includes a robust geospa-
tial component.
INSPIREThe "Infrastructure for Spatial Information in
Europe" (INSPIRE) initiative aims at deliver-
ing integrated spatial information services to
the largest possible number of users. Such
services include visualisation of information
layers, overlay of information from different
sources, spatial and temporal analysis, and
many others. The target users of INSPIRE
include policy-makers, planners and man-
agers at European, national and local levels
and also citizens and their organisations.
To support INSPIRE, OGC and OGCE, the
European OGC subsidiary, provided input to
the early Position Papers for INSPIRE and the
Extended Impact Assessment, evaluating the
political and socio-economic impacts of
INSPIRE. European OGC industry members
participated in the Internet consultation prior
to the adoption of the Framework Directive
by the European Commission. OGCE was
also a partner in the Geographic Information
Network In Europe (GINIE) Project. The aim
of this project was to develop a cohesive
Geographic Information Strategy at the
European level, and which thus provided
strategic input to INSPIRE.
Alessandro Annoni, ESDI Action Leader,
Spatial Data Infrastructures Unit at the Joint
Research Centre (JRC), says: "The adoption
of the INSPIRE directive is expected in 2006.
Non-EU countries including Norway,
Switzerland and Croatia are also likely to
align their national initiatives to the INSPIRE
framework. This means that across Europe
there will be an increasing demand for GI
services based on open standards and forstandards-based portals that will provide
access to multiple "aggregated" sources of
data and online geospatial services. The
development of INSPIRE-compliant infrastruc-
tures initially focusing on the environmental
sector is expected to extend to other the-
matic sectors in near future, from the begin-
ning to include the regional level and also to
influence the local level in some cases."
GMES
The Global Monitoring for Environment andSecurity (GMES), a joint initiative of the
European Commission and the European
Space Agency (ESA), represents a concerted
effort to bring data and information
March 200612
Art i c le
Diagrammatic View of the INSPIRE Vision (from the INSPIRE Architecture and Standards Position
Paper www.ec-gis.org/inspire/).
-
7/27/2019 geoinformatics 2006 vol02
7/51
providers together with users. This way they
can better understand each other and make
environmental and security-related informa-
tion available to the people who need it
through enhanced or new services. Networks
of geolocated in-situ, vehicle-mounted, aerial
and satellite-borne sensors will play a role.
After an initial period of preparation (2001-
2003), the implementation phase is on-going
(2004-2008).
The requirements that drive the implementa-
tion of GMES include: openness, federated
architecture, simplicity of architecture, scala-
bility, dependability, user-friendliness, data
security, quality of service, and global ubiquity
of access.
EUSCThe European Union Satellite Centre (EUSC),
located near Madrid, Spain, is an agency of
the Council of the European Union dedicated
to the exploitation and production of infor-
mation derived primarily from the analysis of
earth observation space imagery in support
of decision-making in the areas of theCommon Foreign and Security Policy (CFSP),
especially the European Security and Defense
Policy (ESDP).
In October 2005, after several interoperability
studies, the EUSC awarded a contract for the
design and support of the EUSC Reference
ture required by the forthcoming INSPIREimplementing rules.
Other EU ProjectsOther EU Projects supporting INSPIRE, GMES
and CFSP/ESDP are:
Open ARCHitEcture and Spatial Data
InfrasTRucture for Risk MAnagement
(ORCHESTRA): September 2004 - August
2007. This Integrated Project is funded by
the EC under the Sixth Framework
Programme that incorporates emerging
specifications of INSPIRE and GMES andcontributes to these initiatives. With the
goal of improving the efficiency in dealing
with risks by enabling interoperability,
ORCHESTRA participants are designing an
open service-oriented architecture and spa-
tial and non-spatial services for risk man-
agement, and are contributing to software
standards for risk management;
Reference Information Specifications for
Europe (RISE): September 2005 - August
2007. This Specific Support Action is funded
by the EC under the Sixth Framework
Programme, also to support INSPIRE and
GMES. Its objectives are data harmonisation
by defining a repeatable methodology and
producing guidelines for the creation of
geospatial data specifications. The RISE pro-
ject focuses on the hydrology and elevation
themes and thus there is a strong linkage
to different GMES applications and the
Water Framework Directive (WFD);
Closely coordinated with RISE is Marine
Overlays on Topography for Annex II
Valuation and Exploitation (MOTIIVE):
September 2005 - August 2007. The focusof this project is on elevation and dynamic
marine features. MOTIIVE is working closely
with the national and international meteoro-
logical and oceanographic communities and
has a key aim to deliver a reference imple-
mentation for a feature catalogue registry;
GEOdata and CRisis Early Warning Situation
Awareness Architecture Concept (GEOCREW):
January 2005 - December 2005. This was a
Specific Support Action funded by the EC
Preparatory Action on Security Research, to
support CFSP/ESDP and GMES. GEOCREW'sobjectives were to integrate various infor-
mation resources for a crisis early warning
"situation awareness" architecture.
Facility (EUSC-RF). A prototype has been
made available for review and testing by
EUSC and other agencies whose operations
mesh with EUSCs.
The EUSC-RF is the result of rigorous analy-
sis of EUSC operations and data sharing
requirements and consideration of the waysin which modern Internet technologies can
support EUSC workflows.
ESA's HMA ProjectIn September 2005, the ESA launched the
Heterogeneous Mission Accessibility -
Interoperability (HMA-I) Project. HMA-I will
define the necessary interfaces and a gener-
ic, service-oriented architecture to ensure
interoperability within the GMES Space com-
ponent comprising a constellation of satel-
lites together with its Ground Segment and
the interfaces to the other components of
GMES.
Besides ESA missions like Envisat and ERS-2,
national missions like Cosmo from ASI (Italy),
Pleiades from CNES (France), RADARSAT-2
from CSA (Canada) and TerraSAR-x from DLR
(Germany) will be providing the Earth obser-
vation data necessary to allow an opera-
tional rollout of the GMES services starting
from 2008. EUSC is also participating in this
joint effort on interoperability and service-ori-
ented architectures based on open stan-
dards. JRCs INSPIRE Team and ESA are work-ing together to ensure coherence of INSPIRE
and GMES developments. According to Pier
Giorgio Marchetti, the HMA Project Manager
at ESA, HMA will leverage recent advances
in interoperability specifications and service
oriented architectures. This will empower the
GMES services with seamless access to
space data from the already identified mis-
sions. The others which will be contributing
in the near future, as well as the necessary
interoperability with the geospatial infrastruc-
March 2006Latest News? Visit www.geoinformatics.com 13
Art i c le
rd Geo-Interoperabilityrd Geo-Interoperability
The GMES capacity as represented by the GMESdiamond, from the Communication from the
Commission to the European Parliament and the
Council, "Global Monitoring for Environment and
Security (GMES): Establishing a GMES capacity by 2008 -
(Action Plan (2004-2008))".
From a presentation at the ESA EUSC IIM workshop
2005, "Web services integration in an European geospa-
tial agency - The EUSC case," by Lucio Colaiacomo
[earth.esa.int/rtd/Events/ESA-
EUSC_2005/Pr05_Colaiacomo.ppt ].
-
7/27/2019 geoinformatics 2006 vol02
8/51
DGIWG and NATOIn September 2005, OGC announced the
signing of a formal Memorandum of
Understanding (MOU) between the Digital
Geospatial Information Working Group
(DGIWG) Secretariat and the OGC. DGIWG,
established in 1983, is a standardization
body comprised of 18 nations whose first
objective was to set up an effectiveexchange standard allowing for the transfer
of data between the countries' geospatial
production agencies, to meet NATO geospa-
tial requirements. A current objective is to
help develop and to encourage the use of
technical standards that enable improved
interoperability and integration of geospatial
information systems. DGIWG wishes to tran-
sition from creating military specific stan-
dards to using consensus industry and inter-
national standards wherever possible.
Brigadier Nick Rigby, Director DGIWG: "Thecollaboration with OGC is a major part of
DGIWG's strategy to work with key partners
in order to maximize the output from our
mutual efforts to deliver geospatial stan-
dards. Such cooperative effort is essential in
the prevailing climate of increasing require-
ments but limited resources."
A recent NATO "information for bid" (IFB CO-
11424-GIS) call for "a common spatial, reli-
able, and time reference framework akin to
enabling the fighting off the same map at
the strategic, operational, and tactical lev-
els." The description of work specifically
mentions OGC implementation specifications,
ISO Geospatial (191XX) standards, and/or
NATO's/DGIWG's ISO specifications (profiles,
application schemas, data dictionaries, etc.)
SpainThe Spanish NSDI, named IDEE for
Infraestructura de Datos Espaciales de
Espaa (www.idee.es) connects geospatial
catalogues, gazetteers and services distribut-
ed in servers all around Spain at different
levels of detail. IDEE is providing servicesthat implement many of OGCs OpenGIS
Specifications. ISO19100 standards, the
INSPIRE initiative and OGC specifications
make possible a framework that has been
the basis of a Spanish Core Metadata stan-
dard (NEM for Ncleo Espaol de
Metadatos), a minimum set of ISO19115
metadata items; a Spanish Gazetteer Model
(MNE for Modelo de Nomencltor de
Espaa); a common schema for Gazetteers
compliant with ISO19112; and a Spanish Data
Model for geographic vector data (MDE forModelo de Datos Espaol) based on ISO
19107 Spatial Schema.
There is a well-established regional SDI in
Catalonia, Infraestructura de Datos Espaciales
de Catalunya, (IDEC). Other Regional (BasqueCountry, Navarra, La Rioja and Galicia) and
Local (Zaragoza) SDIs, National Agency
Reference Nodes and other projects are in
progress. A project called GeoPista has
produced a standards-based local
Government GIS that will be extended to the
more than 4,000 local authorities.
According to Sebastin Mas of the Spanish
NDSI Working Group at the Instituto
Geogrfico Nacional, "The Spanish NSDI
(IDEE) is a powerful emerging reality based
on strong technical support and flourishing
national and regional initiatives proceeding
in a spirit of coordination, collaboration and
sharing of knowledge."
UKThe Ordnance Survey has been licensing dig-
ital Geographical Information (GI) to cus-
tomers since the early 1990s, a business
worth tens of millions of euros. Technical
standards developed in the OGC with the
Survey's active participation have enabled
the Ordnance Survey to offer new capabili-
ties and new benefits to their users. Mostnotably, the Ordnance Survey has provided
the OS MasterMap products, which are dis-
tributed using GML.
This progress has shown the need for further
development of OGC Web Services specifica-
tions. Increasingly, users of geographical
information are using geospatial data to sup-
port decision making processes that involve
automated transactions. In the insurance
industry, for example, or in Location Based
Services (LBS) delivered on mobile devices,
there is a requirement for structured intelli-gent information rich in attribution and capa-
ble of being linked to associated information
stored elsewhere. This often involves chain-
ing of Web services, and not all of the
required open interfaces for these serviceshave been developed into standards.
One of these gaps in being addressed in
OGC's "GeoDRM," or Geospatial Digital
Rights Management initiative. The Ordnance
Survey is leading the development in OGC of
digital rights management techniques. Ed
Parsons, Chief Technology Officer of the
Ordnance Survey, says: "Web Services that
utilize OGC standards will play a crucial role
over the next five years in delivering up-to-
date geographically based decision-making
support to a wide range of users without the
massive management overhead of the past."
GermanyGerman companies and agencies have been
active in OGC for a decade, and they have
implemented standards-based systems at
all levels of government. In 1998 the
Federal Cabinet founded the Interministerial
Committee for Geo Information (IMAGI) and
issued the instruction to develop an effi-
cient geodata management framework for
the federal government. Since then the
Geodata Infrastructure for Germany (GDI-DE) has grown rapidly. In 2003 the chief
executive officers of the Federal
Government and the 16 states took the
decision to implement the GDI-DE in a
combined effort at all levels of public sec-
tor. The GeoPortal.Bund portal which went
on-line during the INTERGEO 2005 confer-
ence in Dsseldorf is to be the central
entry point to the GDI-DE.
The Geodata Infrastructure of North Rhine-
Westphalia (GDI-NRW) was founded in 1999
by the NRW state government, organizingthe efforts and resources of more than 100
institutions. Also in North Rhine-Westphalia,
Conterra implemented the TIM-online
GeoPortal.
March 200614
Art i c le
The five levels of detail defined by CityGML [from "CityGML Interoperable Access to 3D City Models",
Thomas H. Kolbe, Gerhard Grger, Lutz Plmer].
-
7/27/2019 geoinformatics 2006 vol02
9/51
"CityGML," a GML3 Application Profile for vir-
tual 3D city models, is providing important
input into the OGC effort to integrate
geospatial information with information
about built structures. CityGML was devel-
oped by the Special Interest Group 3D (SIG
3D) of the GDI-NRW, chaired by members of
the University of Bonn (Thomas Kolbe and
Gerd Groeger). Many cities are constructing
3D city models for applications such as
urban planning, disaster management,
tourism, vehicle and pedestrian navigation,
facility management, but there is no appro-
priate standard for data exchange. This is
what CityGML developers and others in OGC
are working on.
The NetherlandsIn 2003 the Dutch Directorate for Public
Works and Water Management,
Rijkswaterstaat (RWS), which is responsible
for the maintenance of dikes, roads, bridgesand the navigability of canals in the
Netherlands, started the GeoServices project,
a system for web based access to geo-
information within Rijkswaterstaat using ISO
and OGC standards. The goal was to make
geoinformation available directly from multi-
ple local sources. The use of open standards
was a given, whereas the use of open
source software to realize the OGC web
architecture was one possible choice.
Under supervision of Rijkswaterstaat, the
company Geodan built the GeoServicesapplication architecture and then delivered a
proof of concept, on the basis of which full
implementation was completed.
In 2005 the RWS SDI infrastructure was
access to many more sources of georefer-
enced data, including data from other organi-
zations' OGC-compliant servers on the Web,
and including access to analytical, computa-
tional and data processing facilities. BRGM
represents the Association of European
Geological Surveys (EuroGeoSurveys) within
INSPIRE.
NorwayThe Norwegian Mapping Authority (NMA, or
Statens Kartverk) was identified by the
Norwegian Ministry of Trade and Industry as
a key agency for the success of the
Norwegian e-Gov initiative. A national
"GeoNorge" project was initiated, led by Olaf
Ostensen, Chairman of the ISO/TC 211
Committee and a key player in the harmo-
nization of OGC and ISO specifications.
The first version of the project connected the
servers from National Mapping Agency,Norwegian Geological Survey, Norwegian
Institute for Land Inventory, Public Roads
Administration, and the Directorate for
Nature Management and County
Administration. All servers are made accessi-
ble through their implementations of the
same standard OpenGIS interface specifica-
tions. This was accomplished smoothly and
quickly, without migration of any data, tech-
nical re-engineering, or purchase of new GIS
Servers.
ConclusionEurope's leadership role in the development
of geospatial interoperability standards is
now paying off in the widespread implemen-
tation of these standards. Instead of down-
loading or mailing massive data files, data
sharing is being accomplished with a few
keystrokes in Web-based applications.
Remote online geoprocessing services can be
invoked to derive information, without the
need for the user to own expensive soft-
ware. If progress toward this goal has
seemed slow at times, it now seems to bemoving forward at a remarkable pace.
Martin Klopfer ([email protected]) is the
European Programmes Director of the Open
Geospatial Consortium (Europe) Limited. Guenther
Pichler ([email protected]) is Director
Business Development Open Geospatial Consortium
(Europe) Limited.
expanded with IONIC's standards-based
Redspider products in order to provide more
scalability for transactional and mobile
geo-applications. It is expected that in the
end the RWS Spatial Data Infrastructure will
consist of more than 40 connected applica-
tions.
FranceBRGM is a French public institution with a
mission to acquire, manage and diffuse geo-
logical data necessary for implementing pub-
lic policies relating to the development and
sustainable management of natural subsur-
face resources. To support this mission, in
1998 BRGM created InfoTerre, an Internet site
providing access to geological maps and
other georeferenced data of France.
Soon after InfoTerre started, BRGM adopted a
policy of interoperability concerning the infor-
mation domains for which it is responsible.
In 2001 this resulted in BRGM becoming thefirst supplier of French data to join the Open
GIS Consortium, Inc. (OGC).
The standards-based version of InfoTerre was
released in March 2003, providing users with
March 2006Latest News? Visit www.geoinformatics.com 15
Art i c le
The InfoTerre web site[http://infoterre.brgm.fr/ionic/banques.asp?
mode=expert].
The GeoNorge portal www.geonorge.no/.
-
7/27/2019 geoinformatics 2006 vol02
10/51March 2006Latest News? Visit www.geoinformatics.com 17
European e-GovernmentAward for Dutch Cadastre1.8 Million Requests per Month
European e-GovernmentAward for Dutch Cadastre
Every two years the European Commission presents the e-Europe Awards for e-Government.
The Dutch Cadastre has won the award for the theme Service use: Citizens and
Businesses. The other three awards were granted to the Social Insurance Institution in
Poland, the Agency of Governmental Management in Denmark and the Revenue
Commisioners in Ireland. Arco Groothedde, executive board member of the Dutch Cadastre,
shares his views with us.
By Robin Wevers
Four Different ThemesThe aim of the e-Europe awards is to promote
best practices within the EU so that the corre-
sponding know-how can subsequently be
shared. The e-Europe Awards are organized by
the EIPA, the European Institute of Public
Administration. In 2005 they received 234 sub-
missions for four different themes: Enabling e-
Government, Transformation, Businesses and
Citizens, and Impact. The European
Commission has granted the 'Good PracticeLabel to 76 projects, 52 have been selected
as finalists. The finalists presented their appli-
cations at the e-Government Conference in
November 2005 in Manchester.
Kadaster-on-lineUntil the turn of the century the Dutch
Cadastre had a specific type of software that
customers used to order and retrieve informa-
tion from. This application was not based on
Internet-technology, relatively complex to use
and built on stand-alone systems. It did not
live up to the demands of our time. In 2001
the Dutch Cadastre introduced Kadaster-on-
line. This new application offers up-to-date
information, is immediately available and pro-
vides a user-friendly Internet-interface.
Initially Kadaster-on-line was built for house-
brokers and notaries. These professional cus-
tomers have subscriptions that allow them to
selectively retrieve the required information.
Groothedde: Innovative may not be the first
word coming to mind when hearing the name
Cadastre, but when asked most of our cus-
tomers will undoubtedly qualify us as innova-tive. From 2003 onwards private citizens also
have access to Kadaster-on-line. Groothedde
says: It is the Cadastres task to make infor-
mation available easily and at limited costs.
The electronic highway is an important means
for us to bring real estate-information into
peoples homes and offices.
Verdict of the JuryThe jurys short description of Kadaster-on-line
is: The project delivers online access to land
registry products through national up-to-dateland registry information. It has 6,000 (Note:
this should be 12,000) registered clients and
45,000 users, and provides significant value-
adding service extensions.
The jury qualified the project as : A very
good case regarding innovation in the applica-
tion area. Integration and collaboration of dif-
ferent systems and entities is convincingly
elaborated. It offers high potential for a large
group of users and for replication in different
regions and European countries.
Take-offKadaster-on-line has taken off in a way that
even the people involved could never dream of.
In 2001 175 thousand requests for information
were received every month. In 2005 this number
increased to 1.8 million per month. The number
of professional subscribers has grown from
6,000 to 12,000. Considering that most subscrip-
tions apply to more than one user, the actual
number of users is estimated at 45,000. What
makes Kadaster-on-line different from most
applications? Groothedde: It is different due to
the fact that the starting point has not been the
products the Cadastre can offer, but the demand
of the customers. We truly offer what the cus-
tomer wants. To support this statement he
says: From the moment Kadaster-on-line wasintroduced, we have seen an exponential
decrease in the number of information requests
by phone or at the desk. Kadaster-on-line has a
nice front-end, but most impressive must be the
part you cant see: the systems and data behind
the front-end such as cartographic information,
alphanumeric information, realising the quality of
the data, and logging and billing all transactions.
Kadaster-on-line is the spider in the web. Anice
front-end can only work if the back-office sys-
tems are up to their tasks.
Robin Wevers ([email protected]) is a freelance writ-
er of geo-ICT articles. More information can be found
at www.kadaster.nl and
www.egov2005conference.gov.uk.
Interv iew
The e-Government award.
Arco Groothedde: From
the moment Kadaster-on-
line was introduced we
have seen an exponential
decrease in the number of
information requests by
phone or at the desk.
-
7/27/2019 geoinformatics 2006 vol02
11/51
GeoStore.com Final ComponentImage(ry) is Everything
GeoStore.com Final Component
In March 2006 Infoterra Limited, provider of geographic information products and ser-
vices, celebrates its fifth birthday. Dr Andy Wells, director of sales and marketing,
looks back at the companys history and the launch of GeoStore.
By Andy Wells
Web-based ServiceIn order to meet the demand for accurate and
timely geographic information, capable ofadding immediate insight to any situation,
Infoterra launched GeoStore.com. This web
based service, available since October 2005,
provides aerial data across England which can
be viewed, analysed and purchased on-line. As
the only on-line system of its kind to provide
images that have not been compressed in
storage, images are delivered directly to the
end user without having decreased in quality
due to storage or transmission techniques.
Data ChallengeAs part of Europes space company EADSAstrium, Infoterra was able to draw from a 25
year history of commercial earth observation
and geospatial knowledge. Despite this
breadth of knowledge already within the
group, the challenge of storing and distribut-
ing vast amounts of aerial photography and
other geospatial data was immense. However,for many business processes to benefit from
this valuable resource, the need to solve the
technical challenges associated with storing,
managing and delivering such data required a
solution. Begun in 2000, the 2m research
and development programme finally culminat-
ed in the release of GeoStore.com.
With England-wide aerial photography requir-
ing 13 terabytes of storage, overcoming the
problems of loading, management and viewing
aerial imagery, was crucial.
Infoterra invested in technologies that wouldenable aerial images to be rapidly loaded and
easily managed, while still providing rapid
delivery over thin bandwidths, such as the
web and mobile phones. GeoStore utilises a
technology that allows the user to immediately
receive accurate aerial images, irrespective ofhow the user roams, pans or zooms. This
guarantees that the user waits the same brief
amount of time for the delivery of any image.
EnviSATEstablished 10 years ago for the management
of data from ESA satellites ERS1, ERS2 and
EnviSAT, the data hosting centre in
Farnborough is the largest commercial geo-
graphical data hosting facility in the UK with
1,500 terabytes of data available on or near-
line and an archive in excess of 7,000 ter-abytes.
The core of the system is a storage area net-
work (SAN) linked to multiple server configura-
tions. The choice of the SAN was made follow-
ing extensive testing and optimisation using
geographical datasets. This enables multiple
services to link to a single datastore. The sys-
tem is supported by a number of additional
services including a fire suppression system,
flood detection, environmental control, four
levels of building security, a secondary external
generator, two separate lines from separate
bandwidth suppliers and a nearline file
restoration system capable of replacing 10 ter-
abytes of data in less than a day.
Service BackboneThe latest technology is complemented by a
staff with skill sets required for a successful
geospatial Internet service provider (G-ISP).
This centre provides the service backbone for
GeoStore. In the months following the launch
of GeoStore.com the site has delivered hun-
dreds of aerial images to customers through-
out the world. Two additional services willaccompany GeoStore Select, the service initial-
ly launched that allows a precise location to
be identified and displayed by using roam,
pan and zoom functionality.
The two additional services include GeoStore
Explore, which offers the ability to view and
analyse geographic data through a web brows-
er without the need to store a local copy of
the data or for GIS software. This added func-
tionality is suitable for rapid site investigation,
including risk assessments, planning reviews
and environmental enquiries.Now live, GeoStore Direct enables immediate
access to terabytes of data, delivered on-line
through an open standard solution to the
users own geo-information system. By imme-
diately receiving the image, the user has
March 200618
Special
Tower of London and Bridge, London.
-
7/27/2019 geoinformatics 2006 vol02
12/51
instant access, without having to use his own
hard disk space to save and then insert the
images required. Through an OGC WMS com-
pliant interface, environmental reports, proper-
ty reports, and others can be instantly
enhanced with geographic imagery delivered
through GeoStore.com.
ADS40 Digital Pan Broom CameraInfoterra now services more than a dozen
clients through this hosting centre including
members of the insurance, property, mapping,
government and international community. The
company has continued to invest in new tech-
nologies having recently purchased the ADS40
digital pan broom camera. Unlike frame cam-
those needing support from geospatial infor-mation.
Future plans for GeoStore.com are currently
being finalised. However further data layers are
already being planned for loading and dissem-
ination including the recently acquired
London Heights layer comprising a ground
and building height model with vertical accura-
cies better than 15cm.
Dr Andy Wells ([email protected]) is
head of sales and marketing at Infoterra. Have a
look at www.infoterra.comfor general informationon Infoterra. Visit www.GeoStore.com to learn more
about the product discussed in this article.
eras this device is able to capture a seamless
strip of digital imagery over large areas. Using
a three-line stereo sensor, the camera can
simultaneously capture red-green-blue and
colour infra-red imagery, providing the rapid
production of both orthophotos and surface
elevation models. The ADS40 has so far been
used to capture 1,400 square kilometres ofBerkshire, producing 16 lines of imagery that
were mosaiced into a single image within a
week.
ConclusionFor Infoterra, GeoStore.com is the final compo-
nent in a corporate strategy initiated in 2003
to provide a complete end-to-end solution for
March 2006Latest News? Visit www.geoinformatics.com 19
Special
n Corporate Strategyin Corporate Strategy
The challenge of storing and distributing vast amounts of aerial
photography and other geospatial data was immense.
-
7/27/2019 geoinformatics 2006 vol02
13/51
Georeferenced Oblique Aerial IPictometry - Intelligent Visual Technology
Georeferenced Oblique Aerial I
With todays emphasis on visual information, GIS professionals are looking closely at a
third format: georeferenced digital obliques. With this, users are taking geospatial data
into non-traditional areas of utilization and placing it into the hands of those who need
visual information for fast, accurate decision making. The key is the georeferencing
component, which gives the imagery an in-built functionality and presents the user
with a wide range of potential applications. For the Geomatics industry the technology
offers increased information availability with a remarkable operational effectiveness.
by Frank Arts
IntroductionThe development of georeferenced oblique
imaging technology has been spearheaded
by US-based Pictometry International Corp.,
headquartered in Rochester, New York. Its
patented imaging systems became commer-
cially available more than six years ago. The
company has generated an extensive image
library on a county-wide level. These are
comprised of high-resolution, georeferenced
verticals and obliques, and can be utilizedby various government agencies and private
industry. Flown in a grid pattern, the vertical
and oblique images are captured simultane-
ously. Averaging approximately 800 square
miles per county, the coverage is extremely
dense with every square foot on the ground
accounted for. In fact the company has been
contracted to fly 30% of the countrys popu-
lated areas.
Pictometry has taken the complex georefer-
encing dynamic and integrated its functionali-
ty with oblique visual imagery, using
inertial/GPS technology from Canadian-based
Applanix, a company active in the field of
inertial geospatial solutions. The merging ofhigh-resolution digital obliques and accurate
georeferencing has proven to be a potent
combination. Since its launch onto the
Geomatics market, the technology has been
adopted by several organizations for applica-tions as emergency management, land-use
planning, strategic analysis, and homeland
security.
Engineered to produce high-resolution images,
it is an accurate second order visualization
tool, integrated with multi-functional software.
It is not designed to replicate survey-grade
precision, but it does use high-accuracy LiDAR
data and digital elevation models (DEMs) to
produce a pixel-level ground sample distance
(GSD) in the order of 15 centimeter.
Data InteroperabilityThe technology makes it possible to effec-
tively combine digital imagery and vector
data, and to exploit the geo-relationship
between the two data formats. To be truly
useful raw imagery must be geometrically
corrected and geographically encoded to
ensure both vector and image data have a
common registration. Using rectification and
geo-correction techniques on vertical imagery
produces an orthophoto with the image dis-
tortion removed, allowing georeferenced vec-
tor data to be overlaid with true positional
accuracy.
Pictometry is now introducing the technology
into Europe and beyond, with licensing rep-
resentation through Blom ASA in Norway,
Simmons Aerofilms in the UK, and
Compagnia Generale Ripreseaeree (CGR)
Group in Italy. David Critchley, Simmons
Aerofilms Director Aerial Operations and
Blom COO, states: The technology is being
used primarily to capture urban areas. We
have embarked on a project to cover every
town and city with a population greater than50,000, which we estimate to be around 800
targets. Currently, 60 cities in Italy and 12 in
the UK have been completed, and this year
we expect to increase the total cities cap-
tured to around 200, including London,
Amsterdam, Rome and Madrid. We are cap-
turing these cities at 15cm resolution but
have not ruled out undertaking whole coun-
ties similar to the US model.
The primary applications driving interest in
the UK are homeland security and property
taxation. Simmons has just completed anarrangement with Ordnance Survey (GB)
which has total market penetration for GIS
and mapping, giving them sole agency sta-
tus for re-selling Pictometry in Britain.
March 200620
Special
The Coliseum in Rome. Image courtesy of Compagnia Generale Ripreseaeree (CGR), part of the Blom ASA Group.
-
7/27/2019 geoinformatics 2006 vol02
14/51
Application PotentialThe application potential for this type of
visual technology is wide-ranging. The use of
digital imagery as an information source
allows technical and non-technical users
alike to acquire a quick appreciation of the
geospatial relationship between their data
and the real world. Location-specific informa-
tion, utilizing a three-dimensional coordinate
system, becomes its most effective when it
relates to a point on the earths surface.
When the information is displayed as co-
located image and vector data, a GISbecomes intuitive and visually dynamic- a
key to maximizing data comprehension and
ultimately, data utilization.
Pictometry believes that with the develop-
ment of georeferenced oblique imaging and
its integration with GIS technology, the appli-
cation potential for oblique visuals has
expanded.
For many organizations implementing the
technology with their GIS, there have been
substantial benefits, primarily through sav-
ings in time and operational expense.
Detailed analysis of oblique imagery can
yield the same types of information that
would be obtained in the field, without
putting personnel at risk working beside a
busy highway for example, or interrupting
traffic flow while undertaking bridge mea-
surements. Updating or verifying planimetric
detail, or locating and cataloging signage,
forward. Unless you are very familiar with
viewing vertical images, to the un-trained eye,
it is not always obvious what is being seen.
The plan view of buildings can be the most
deceptive, with variations in roof elevation
often the most difficult to detect. However,
from an oblique vantage point, high-rise build-
ings, towers, bridges, retaining walls and other
tall features are easily identified. You can see
the fronts of buildings, and also the sides and
backs. The information on billboards and shop
facades can be read, and very quickly a com-
plete and detailed impression of the area can
be established.
Objects obscured by overhanging vegetation
or located under bridges and overpasses can
now be seen. Ridges, valleys and cliffs, which
illustrate dramatic changes in topography,
become quite distinct. Oblique imagery pro-vides the user with a wealth of detail and
visual information that is just not discernible
on a standard vertical photo.
Unless you are viewing vertical imagery stereo-
scopically, using image pairs with a 60% over-
lapping common area, any form of relief on a
single vertical photo is not readily apparent. It
requires an experienced photogrammetrist
familiar with the three-dimensional aspect to
extract the relevant information.
Enterprise-wide Data SharingWith the attributes associated with georefer-
enced oblique digital imagery, this technolo-
gy has opened the doors to enterprise-wide
data sharing. Dante Pennacchia, Chief
are just some of the tasks that can be car-
ried out quickly using this technology. For
companies involved with transportation plan-
ning or highway asset management, field
trips to validate roadside data have been
reduced by as much as 70%, the majority of
which is in the elimination of travel time get-
ting to and from the job site.
Perspective AdvantagesFor the user, the primary advantage of oblique
imagery is perspective familiarity. The ground
is being seen from a more recognizable angle
and therefore feature identification is straight-
March 2006Latest News? Visit www.geoinformatics.com 21
Special
magerymageryAbout the AuthorFrank Arts has spent more than 20 years in photogramme-
try, technical training, marketing and business development
with various aerial survey and engineering companies in
Toronto. He has worked on several mapping projects
throughout Canada and the United States. Of particular inter-
est have been a number of projects involving forensic analy-
sis and crime scene photogrammetry for law enforcement
and investigation.As a technical instructor he has been responsible for training
staff in digital mapping methodology, primarily in the areas
of air photo interpretation, aerial triangulation and pho-
togrammetric data capture. Through his marketing and business development experience he
has produced a number of technical publications including white papers, business proposals
and promotional material.
Most recently as a technical communications writer with Applanix Corporation, his industry
knowledge was focused on editorials, case studies and marketing collateral dealing with air-
borne, land and marine survey applications utilizing integrated inertial/GPS technology.
Closely involved with photogrammetry through its transition from analogue to analytical, and
to the virtual automation of softcopy technology, Frank has a keen interest in the potential
for new applications. His current interests focus on technologies such as LiDAR, IfSAR andthe integration of multi-sensor imaging systems, and the impact they are having on aerial/ter-
restrial photogrammetry and remote sensing applications.
Frank is based in Canada, and as a technical writer is the North American correspondent for
GeoInformatics magazine. He will be at this years ASPRS conference in Reno should you
wish to meet him.
Police squad car with on-board Pictometry system. Imagery courtesy of Pictometry International Corp.
-
7/27/2019 geoinformatics 2006 vol02
15/51
Marketing Officer at Pictometry explains: We
call it the democratization of GIS informa-
tion, by allowing the technology to expand
into areas which until now had not been
major users of GIS and digital imagery. This
can be seen in the case of policing and
emergency task force response activities. Pre-
emptive and synchronized tactical planning
are prime applications where visual reference
information can play a vital role.
He continues: The system can be tailored
for a specific environment by enhancing the
capabilities of a GIS and placing it in a
patrol vehicle or police cruiser, with mobile,
touch screen functionality. Law enforcement
officers, emergency medical personnel and
first responders, can all access multiple
obliques of the same location at the same
time and very quickly put into place a com-
bined action plan.
Intelligent Images
The system is designed to give the user the
maximum amount of visual information for
the areas they are interested in. County-wide
image data is updated every two years, but
depending on the amount of redevelopment
and physical change taking place, particular
regions can be re-flown as frequently as
every six months. This ensures the imagery
remains as current as possible and is a rele-
vant image source for a wide-range of poten-
tial users.The power associated with this type of visual
information lies in its interactive capability.
Pictometrys multi-functional Electronic Field
Study (EFS) software is an integral part of
the technology which allows the user to very
effectively manage large quantities of visual
information, while importing shapefiles and
re-projecting vector overlays using oblique-
March 200622
Special
Hurricane Katrina change analysis images. Imagery courtesy of Pictometry International Corp.
Advertentie DATEM 185x134 mm 07-12-2005 10:52 Pagina 1
-
7/27/2019 geoinformatics 2006 vol02
16/51
viewing tools.
Using a geo-coded street address or a geo-
located entry in a GIS database, location-
specific features can be pinpointed on a
series of low obliques (taken from different
angles) and imported for detailed analysis
and investigation. Relevant geospatial vector
data can then be introduced and the com-
plete GIS information level expanded with a
task-oriented practicality. The user can obtain
direct length, width and height measure-
ments of any visible feature, together with
perimeter distances, area calculations, coor-
dinate values, compass bearings and eleva-
tions. Horizontal distances can even be gen-
erated in an as the crow flies, or ground
plane surface format, referred to as walk
the earth.
The ability to extract this type of information
from an oblique image, plus the opportunity
to integrate position-accurate vector data
and its attributes, has produced a powerful
planning and analysis tool. An embeddableversion of the software is also available
which allows it to be integrated with various
third-party GIS products, such as ESRIs
ArcGIS and ArcIMS, as well as Autodesks
MapGuide. In addition, the EFS software
enables users to retrieve, overlay and store
GIS data directly from ESRIs ArcSDE.
adjoining buildings, is it a high-risk location,
are there hazardous materials located inside
(garage, chemical storage, paints etc.), is
neighborhood evacuation required? All this
information can be made available to emer-
gency personnel right at the scene by inte-
grating oblique imagery and various types of
GIS data.
At night, or in a smoke-filled environmentwhere on-site visibility is severely restricted,
this type of information becomes even more
vital as emergency crews are able to view
multiple, full colour, daytime oblique views
of the location. This was the situation faced
by the Polk County fire department in 2004,
when visibility at one particular fire scene
was virtually zero. In analyzing the obliques
on the command vehicles laptop, chemical
storage tanks were observed close to the
burning building, which could not be seen
by the fire crew. Immediately, hoses wereredirected to spray the tanks to keep the
embers from igniting them. All the necessary
decision making information was pulled
together very quickly and the safest and
most operationally-effective plan put into
place.
Ultimate Visual ResourceGeoreferenced oblique imaging technology is
dynamically changing the way we see our
world. Today, there are very few geographic
information users who do not employ visual
image data in the decision making process.
Georeferenced obliques can expand that pro-
cess as the ultimate visual resource. Whether
it is for project planning, or disaster
response and recovery, the availability of cur-
rent, georeferenced oblique imagery in the
hands of people who need it the most can
be the determining factor in project success
or mission failure.
AcknowledgementsSpecial thanks to Will Smith, Marketing
Manager and Steve Schultz, Chief TechnologyOfficer, Pictometry International Corp.
Frank Arts ([email protected]) is a
technical writer with particular interest in aerial
and terrestrial photogrammetric applications for
the Geomatics industry. Have a look at
www.pictometry.com to learn more about the
technique discussed in this article.
Currently, EFS is undergoing integration with
MapInfos product suite.
Faster Decision MakingFor emergency services such as a fire depart-
ment, the system has proved invaluable. In
Gwinnett County the technology is already
available on 18 of its 83 vehicles. The entire
fleet will eventually be equipped enabling all
fire crews to access the system enroute to a
fire anywhere in the county. Vertical images
and obliques can be analyzed to establish
the best locations to park fire trucks and
position hoses. Laneways and gates etc. can
be quickly measured to determine whether
access is possible. Doors, ramps and win-
dows can be identified and evaluated as
potential access and escape routes.
The heights of buildings can be accurately
calculated to verify ladder reach and water-
cannon elevation, and an effective and effi-
cient action plan formulated before the crew
arrives at the scene of the fire.By importing building plans and integrating
GIS data to locate fire hydrants and water
supply lines, a very detailed description of
the building and the surrounding area can
be assembled. Is it a multi-storey building, a
residential or commercial unit, does it have a
flat or peaked roof, what is its proximity to
March 2006Latest News? Visit www.geoinformatics.com 23
Special
GIS flood plain overlay. Imagery courtesy of Pictometry International Corp.
With georeferenced oblique imaging technology users are taking
geospatial data into non-traditional areas of utilization and placing it
squarely into the hands of those who need visual information for fast,
accurate decision making.
-
7/27/2019 geoinformatics 2006 vol02
17/51
In the Gigapixel article in GeoInformatics 1-2006, the various methods that have been
devised to generate Gigapixel frame images in the area of amateur and professional
photography have been outlined. In this article, a review will be undertaken of the
methods being used by scientific astronomers to generate large-format digital frame
images that are Gigapixel in size. These appear to have much more relevance to
satisfy the current needs of the photogrammetric and remote sensing communities
for the direct acquisition of larger-format frame images from airborne platforms.
by Gordon Petrie
Astronomical ImagesUntil recently, most of the images collected by
astronomical telescopes have been recorded
as frame images on large-format photographic
plates. Thus, for example, the wide-angle
Schmidt telescopes that have been used to
conduct systematic surveys of the northern
skies [by the Palomar Observatory in the
U.S.A.] and of the southern skies [by the U.K.Schmidt Telescope (UKST) based in Australia
and the European Southern Observatory (ESO)
located in Chile], record their images typically
on 14 x 14 inch (35.5 x 35.5 cm) photographic
plates. The images on these photographic
plates are then digitized using automated
plate scanners such as the SuperCOSMOS at
the Royal Observatory Edinburgh (ROE). This
uses a 10 m pixel size and has a measuringaccuracy of less than 1m. Thus it operates
with pixel sizes and accuracies similar to those
of the current Leica, Z/I Imaging and Vexcel
film scanners that are used in aerial pho-
togrammetry. The size of the final digital image
is 35.5k x 35.5k pixels = 1.26 Gigapixels.
However, in recent years, in scientific astrono-
my, as in aerial photogrammetry, the trend has
been towards the direct acquisition of digital
images. So existing telescopes are being con-
verted and new telescopes are being built to
allow the direct digital imaging and recordingof star images.
As already discussed in the previous
Gigapixel article, a major difficulty, both for
photogrammetrists and astronomers, is thelimitation in the size of the area arrays that
are currently available for direct imaging.
Besides which, the CCD arrays used in astro-
nomical telescopes need to detect the very
low light levels that come from very distant
stars. So they need to have high quantum
efficiencies and very low noise levels. As a
result, many of the CCD arrays installed in
astronomical telescopes are custom built by
specialist labs. Furthermore, the yield that
is the number of CCD chips that reach the
required level of performance - is often low,which makes them rather rare and very
expensive. However, in order to overcome
the problem of the small size of the individ-
ual area arrays, astronomers are now devel-
oping mosaics of these area arrays that will
generate large-format digital images that will
replace their photographic plates. In particu-
lar, their recent development of buttable CCD
arrays could be of considerable interest to
the photogrammetric and remote sensing
communities.
1 Ground-Based TelescopesExamples of ground-based astronomical tele-
scopes where CCD area arrays are being
used to produce large-format frame images
include the Canada-France-Hawaii Telescope
(CFHT) and the Sloan Digital Sky Survey
(SDSS).
(a) Canada-France-Hawaii Telescope (CFHT)
This telescope is, as its name suggests, a
joint project between Canada's National
Research Council (NRC) and the French
Centre National de la Recherche Scientifique(CNRS) in cooperation with the Institute of
Astronomy of the University of Hawaii. For
many years, it has operated a telescope with
a 3.6m diameter primary mirror from an
observatory located on the summit of Mount
Kea in Hawaii. The use of CCD area arrays
on this telescope has evolved steadily using
a series of digital frame cameras. First it was
equipped with a single Loral 2k x 2k pixel
array in 1991; then a 4k x 4k mosaic
(MOCAM) in 1994; next an 8k x 8k array
(UH8K) in 1995; and finally a 12k x 8k mosa-ic (CFH12K) in 1999. This last mosaic com-
prised twelve 2k x 4k pixel CCD area arrays,
each of which was buttable on three sides,
that were fabricated by the MIT Lincoln
March 200624
Special
Figure 1 (a): The MegaCam digital frame camera
mounted on a cart for transport. The camera was
built by the French Atomic Energy Commission (CEA)
and fitted to the Canada-France-Hawaii-Telescope
(CFHT) which is located in an observatory on the sum-
mit of Mauna Kea, Hawaii. (Source: CEA, France)
Figure 1 (b): The MegaCam CCD mosaic featuring 40
buttable CCD detectors, each 2k x 4k in size, manufac-
tured by the e2v technologies company in the U.K.
(Source: CFHT Corporation)
Gigapixel Frame Images: PartIs the Holy Grail of Airborne Digital Frame Imaging in
Gigapixel Frame Images: Part
-
7/27/2019 geoinformatics 2006 vol02
18/51
Laboratory. These arrays were arranged in
two rows, each consisting of six individual
arrays that produced images that were 12k x
8k = 96 Megapixels in size.
In 2003, a further development saw the
introduction of a still larger mosaic on the
so-called MegaCam that was built by the
French Atomic Energy Commission (CEA) andfitted to the telescope, see Figure 1(a). This
utilizes 36 (later 40) arrays, each 2k x 4k
pixels in size, producing a total image size of
340 (later 377) Megapixels , see Figure 1(b).
These gaps are eliminated by a dithering
technique that is implemented as the tele-
scope tracks the sky field. However this is
not a observing procedure that seemsapplicable to imaging from airborne plat-
forms. Instead some other arrangement
needs to be found to fill the tiny gaps.
The basis for this might well be similar to
that adopted for the Vexcel UltraCam air-
borne digital frame camera.
(b) Sloan Digital Sky Survey (SDSS)
This project follows on from the systematic
sky surveys using wide-field Schmidt tele-
scopes described above. It aims to cover
only one-quarter of the entire sky (in theNorthern Hemisphere) but in much greater
detail - since its CCD detectors can resolve
much fainter objects than could be detected
using photographic plates. The Sloan Survey
The VST telescope of the European Southern
Observatory (ESO) and the MegaCam fitted
to the MMT telescope of the Smithsonian
Astrophysical Observatory (SAO) will use sim-
ilar sets of 36 close-packed CCD arrays in
their focal planes. These new sets of four-
sided buttable CCD arrays have all been
supplied by the e2v company from the U.K.It should be noted that, although these
arrays are fully buttable, there still remain
very small gaps (of several tens, up to one
or two hundred pixels) between the arrays.
March 2006Latest News? Visit www.geoinformatics.com 25
Special
Figure 2 (a): The telescope of the Sloan Digital Sky Survey (SDSS) forms part of the Apache Point Observatory located in
the Sacramento Mountains of New Mexico. The telescope has a 2.5m diameter primary mirror. The boxy slatted metal
structure is the outer wind baffle, mounted separately from the rest of the telescope, to help prevent the wind from
shaking the telescope.
Figure 2 (b): The 30 CCD area arrays fitted to the
SDSS telescope are arranged in 6 columns, each con-
taining 5 arrays. Each column is encased in a vacu-
um sealed chamber, while each CCD in an individual
column has a different spectral filter placed in front
of it in order to generate a false-colour image.
(Source: Sloan Digital Sky Survey)
Iight?
I
-
7/27/2019 geoinformatics 2006 vol02
19/51
is based on the use of a dedicated 2.5m
diameter telescope with a wide field of view
that is located at the Apache Point
Observatory in New Mexico, see Figure 2(a).
This site is 2,800m in elevation and well
away from any city or town that might gen-
erate extraneous light that would affect the
exposure of the sky image. The SDSS tele-
scope is equipped with 30 CCD area arrays,
each 2k x 2k = 4 Megapixels in size manu-
factured by the Scientific Imaging
Technologies company based in Oregon.
These are arranged in 6 columns, each con-
taining five arrays, see Figure 2(b). The tele-
ground-based wide-field telescopes that can
undertake systematic surveys of the sky with
a view to detecting and locating ever fainter
objects. These are the PanSTARRS
(Panoramic Survey Telescope & Rapid
Response System) and the LSST (Large-aper-
ture Synoptic Survey Telescope). Both are
based on the use of very large mosaics of
CCD area arrays that will produce Gigapixelimages.
(c) PanSTARRS
This particular project is being undertaken by
the Institute of Astronomy of the University
of Hawaii. It is designed specifically to detect
potentially hazardous objects in the Solar
System (primarily asteroids) that might hit
the Earth. Their detection will be achieved
through repetitive observations of the same
piece of the sky over time scales of a few
days or weeks. However the wide field ofview of its optical system makes it suitable
for other types of astronomical project. The
PanSTARRS observing system will comprise
an array of four telescopes, each equipped
with a 1.8m diameter primary mirror and its
own mosaic of CCD area arrays. Normally
these four telescopes will operate together
and point in unison at a particular part of
the sky under computer control.
The use of four relatively small but wide-
angle telescopes is thought to be cheaper
and quicker to build than a single telescope
of a comparable light gathering area. The
CCD arrays to be installed in the focal plane
of each telescope are being developed by
the MIT Lincoln Laboratory. Each will consist
of an 8 x 8 mosaic of orthogonal transfer
arrays (OTAs). In turn, each of these OTAs
comprises an 8 x 8 mosaic of orthogonal
transfer CCD area arrays, each 512 x 512 pix-
els in size. Thus, in total, each camera will
record images with (8 x 8) x (8 x 8) x (512 x
512) pixels = one Gigapixel in size, see
Figure 3(a). Financing such a large project as
PanSTARRS has proven to be q