CC BY
80
OPENPHOTOGRAMMETRY - КОНЦЕПЦИЯ СТРУКТУРЫ ОТКРЫТОЙ СИСТЕМЫ ОБМЕНА РЕЗУЛЬТАТАМИ ИССЛЕДОВАНИЙ С ОТКРЫТЫМ ИСХОДНЫМ КОДОМ
Евгений Левин
Мичиганский технологический университет, Институт технологии, 1400 Townsend drive, Хоутон MI 49931, США, зав. кафедрой прикладной геодезии, доктор наук, сертифицированный фотограмметрист, тел.+1 (906)487-2446, e-mail: elevin@mtu.edu
Гай Хемброфф
Мичиганский технологический университет, Институт технологии, 1400 Townsend drive, Хоутон MI 49931, США, Хоутон, США, зав. кафедрой администрирования компьютерных сетевых систем, профессор, тел.+1 (906)487-3248, e-mail: hembroff@mtu.edu
В статье описываются структурные и организационные принципы веб-портала для обмена и аппробации фотограмметрических моделей сенсоров. Ученые фотограмметристы всего мира смогут обмениваться результатами своих работ в форме методов и алгоритмов с помощью описываемого портала. Операции системы будут также иметь академические применения в геопространственном образовании.
Ключевые слова: фотограмметрия, модели сенсоров, системная совместимость, киберинфраструктуры.
OPENPHOTOGRAMMETRY - CONCEPT OF OPEN-SOURCE FREE COLLABORATIVE FRAMEWORK
Euegene Levin
Michigan Technological University, School of Technology, 1400 Townsend drive, Houghton MI 49931, Chair of Surveying Engineering, Doctor, Certificated Photogrammetrist, tel. +1 (906)487-2446, e-mail: elevin@mtu.edu
Guy Hembroff
Michigan Technological University, School of Technology, 1400 Townsend drive, Houghton MI 49931, Chair of Computer Network System Administration, Associate Professor, tel. +1 (906)487-3248, e-mail: hembroff@mtu.edu
The paper describes a design concept of framework and web-portal for photogrammetric sensor models exchange and evaluation. Photogrammetric imaging scientists worldwide will be able to share results in form of methods and algorithms via OpenPhotogrammetry. Activity will also impact geospatial education.
Key words: Photogrammetry, Sensor Modeling, Interoperabilty, Cyberinfrstructures. SUMMARY
In recent years, the field of Geographical Information Sciences has greatly benefited from the dramatic increase in amount and availability of high-quality geospatial imagery. However, remotely sensed imagery cannot be immediately integrated into cyberinfrastructures. This is due to fact that raw imagery obtained from satellites, UAVs, and non-metric terrestrial cameras needs to be processed
photogrammetrically before further deployment. Typically photogrammetric processing encompasses: sensor calibration, georeferncing based on rigorous sensor model or rational polynomials, triangualtiuon with ground control densification, rectification and/or orthorectification, 3D feature and DEM extraction; and requires use of specialized commercial software licenses. Online photogrammetric processing frameworks do not exist. Current paper describes a structure of free open-source worldwide-opened consortia and geospatial web-portal challenging photogrammetric processing of remotely sensed imagery.
1. INTRODUCTION
Need in timely and interoperable photogrammetric image processing is recognized by geospatial science and technology. Some influential research papers [Clark et al, 2005], [Lillesand et al, 2000] recognizes need for changes in geospatial information science and technology including needs in open-source interoperable research frameworks [Longley 2002]. This indicates that increased geospatial tools interoperability will help in geography and other humanity research success.
The most fundamental approach to the development of open and free photogrammetric applications for students and researchers is maintained by ISPRS Computer Assisted Teaching Contest group [CATCON]. Few of these systems are available for online use, like [OpenDragon] and Learning Digital Photogrammetry [LDP]. However, those systems a) are mostly oriented for teaching and learning rather then research, and thus deploy a fixed set of test-images without the possibility to upload new images; and b) deals mostly with 3D point-clouds generation and remote sensing then with photogrammetric sensor modeling. Image-independent functionality, but in standalone mode, is provided by Insight3D open-source free system [Insight3D]. GRASS GIS (Geographic Resources Analysis Support System) freeware with powerful remote sensing functionality is used by many academic institutions worldwide [GRASS].
Very interesting web-based project for close-range cases entitled Architectural Photogrammetric Network tools for education [Arpenteur] has multiple capabilities. It was developed deploying the Java Development Kit (JDK) and based on the JAVA virtual machine 5.0 and Java Advanced Imaging (JAI) library. Unfortunately, project has been inactive since 2008. Google Building Maker for Google Sketchup [GBM] represents a very useful example for integration of some photogrammertic data obtaining functionality into open geospatial system.
The most useful for further developments open-source resources include: Python Photogrammetry Toolbox [PHT], Stereo Photogrammetry Extractor [SPE] and Basic Image Algorithms library [BIAS].
On cyberinfrastructures [NEC] ,[NEC] side one of the most recent NASA collaborative network developments is a NASA Earth Observing System (EOS) Higher-Education Alliance (NEHEA) GeoBrain project [Chen et al 2011], [Deng and Di, 2010] ,[Di 2003]. Development of open-source web-portal can be integrated and be considered as a data obtaining service for such cyberinfrustructures. The challenges of “OpenPhotogrammetry” are:
- Accelerate deployment of remotely sensed geospatial datasets in GIS and virtual 3D globes driven decision support system by free photogrammetric processing services;
- Enable access to photogrammetric imaging research for educational communities and students from many parts of the world where average price around $10,000 [Petrie ]per photogrammetric software license is not affordable;
- Simplify communication and collaboration of photogrammetry researchers by deployment of ubiquitous network services such as LinkedIn, Facebook, Google+, etc.
2. PHOTOGRAMMETRIC SENSOR MODELS AND THEIR INTEROPERABLE DESCRIPTION PRINCIPLES
Deploying rigorous photogrammetric sensor modeling allows extracting the most accurate metrics from real-world geospatial images. For the establishing of “OpenPhotogrammetry” the following geospatial imagery types should be considered in general:
1. Images of metric cameras in central projection, allowing carrying out of standard interior orientation procedure with determination of principal point position, known principal focal distance, and sometimes known optical system distortions. That imagery may be acquired from terrestrial, manned/unmanned aerial and satellite platforms.
2. Images of non-metric cameras in the central projection, with unknown interior orientation parameters. Presently the digital still cameras of such non-metric type are widely distributed, especially from small-UAV.
3. The images received by various scanning systems, including space based. Mostly such imagery includes high-resolution satellite observation systems, such as Landsat, Ikonos, QuickBird, Spot etc, and also aerial scanning system such as Leica ADS line scanner
4. Geospatial imagery obtained as result of active scanning system (LIDAR, SAR, InSAR) data processing. Scanning point-clouds and data are obtained from terrestrial, aerial and satellite platforms.
Typical geometries of geospatial imagery scanners are shown on Fig. 1.
Figure 1 Geospatial scanning systems geometries and physical principles [Sabins,
1987]
“OpenPhotogrammetry” will operate objects representing different types of solutions for various sensors listed above. Sensor models for solutions can be separated in two categories: rigorous sensor models operating physical sensor parameters and generic mathematical models. The most popular of generic mathematical models is a rational function model [Di et al 2003]. A rational function is defined as a polynomial divided by a polynomial. Functions can be defined for algorithm fitting, with the independent variables being 1, 2, or 3 -dimensional. The power of the rational functions can vary for each dimension as well as for the numerator and the denominator. Space Imaging Inc. and Digital Globe Inc. are replacing rigorous model parameters by rational polynomials coefficients (RPC) for IKONOS and QuickBird satellite imagery respectively. However, due to [Toutin,2000], [Grodecki 2001] processing scanning-type imagery such as IKONOS by means of rational function polynomials in practice produces results 4-5 times less accurate than processing based on rigorous geometrical models. Classical central projection model deploying six parameters of 3 projection coordinates and 3 rotation angles will work well for traditional metric cameras, but will not provide as good of results as for scanning type imagery. Research of sensor models which work for scanning type imagery and provide an accurate result is one of the major topics in today’s photogrammetric sensor modelling research. Open Geospatial Consortia [OGC] in compliance with Federal Geospatial Data Committee [FGDC] established multiple interoperable standards including the most important, for
“OpenPhotogrammetry” sensor modeling markup language -SensorML [Botts 2005], [Chen et al, 2011].To achieve mentioned above capability interoperable, incremental
and OGC-compliant photogrammetric imaging and sensors descriptions. These descriptions should allow geospatial researchers to modify existing sensor models and include their own sensor models into “OpenPhotogrammetry” Fig. 2 outlines our initial XML-like object-oriented approach to photogrammetric sensor models descriptors.
< image G-Sat>
< SensorML Compliant data>
Vendor, Owner,Date Uploaded,Time,Date, Auxiliary data
< SensorML Compliant data>
<sensor Model>
<Author>
"Student Johnson "
<Author>
<Model Name>
‘‘Affine transformation "
<Model Name>
<Number and Names of Sensor Parameters>
6, ‘‘a0 ", ”a1", "a2 ", "b0 ", "b1", "b2 "
<Number and Names of Sensor Parameters> <ModelMath>
‘‘Raw=a0+a1 *X+a2*Y;
‘‘Col = b0+b1*X+b2*Y;
<ModelMath>
<sensor Model>
* * *
< image G-Sat>
Figure 2. Sample of approach to photogrammetric sensor models interoperable
description.
It is visible from Fig. 2, that the record for any sensor may support several sensor models developed by different researchers. Object-oriented structure is very important for establishment in the future knowledge base of sensors and sensor models that will be opened for queries by any free extension of SQL. It is also important to provide efficient parsers of photogrammetric sensor models and to build functionality controlled by an easy-to-use human-machine interface. The purpose of this interface is to allow selection of appropriate sensor model for teaching and/or research purposes with execution of ground control points measurements (if any) and accuracy estimation. Preliminary ideas of human-machine interface design are outlined on Fig. 3. When sensor model parameters are found, “OpenPhotogrammetry” software service will be able to perform image rectification functionality by combining DEM and source image. Given the opportunity to determine the sensor parameters for specific image or satellite scenes, it will be rational to build another functional capability, targeted mostly on geospatial students to manipulate with different views of the same terrain by re-sampling of orthoimage.
Source Image to Upload (selection)
Image OverView
Auxiliary data, Ground Control Points(GCP) X,Y,Z Enter
Select Model Type:
Rigorous
RPC
DLT
Combined
Specific Sensor Computer
Model Selection
Measure GCP on Image
Input/Import data from Camera Calibration Report (if exist)
Fiducials/Reseau Measurements. Interior (if exist)
Accuracy Estimation Scrolling Listing Residuals, RMS, Error Ellipses.
Save
Results
Export
Metadata
Figure 3. Initial Design of Sensor Model Manipulation Man-Machine Interface.
Initially this learning application is foreseen as working as demonstrated on Figure 4.
OrthoImage
View
►
Select DEM File or permanent elevation
Result: Synthetic Image of Selected Sensor Model
Select Sensor Type and Model
Resample
Pictorial Sensor
Save Image and Metadata
Figure 4. Initial design of OpenPhotogrammetry geospatial sensor model learning
application
By means of the learning application, students will be able to generate various synthetic aerial and satellite imagery and learn from hands-on involvement about various sensor models and systems. The research value of the synthetic images and metadata generated by the application is the testing of new algorithms and sensor models. The advantage of synthetic images coupled with metadata is that all their parameters are known and modeled. Use of these images can help in verification of new sensor model methods, algorithms and software by direct comparisons of the results obtained with source data. In general, DEM is needed for learning “OpenPhotogrammetry” tools execution. It can be uploaded from various geospatial data warehouses when available. Otherwise DEM can be extracted from overlapped geospatial images which may require DSM generation mode development.
Multiple research problems have to be resolved in order to proof a feasibility of the outlined “OpenPhotogrammetry” workflow. As for any complex innovation, it is necessary to resolve several contradictory requirements: a) between openness of the system architecture and necessity to avoid breach of imaging vendors copyrights; b) between optimality of the multi-tier architecture and lack of internet connectivity and connection quality compatible with architecture in many developing countries; c) between interoperability and users-capability of photogrammetric algorithms customization; and finally d) the fundamental contradiction known as “training versus education” [Levin et al, 2010] well known in concern to educational use of the “OpenPhotogrammetry” framework. Initial problem solving ideas are: via credible user-authorization system(ACXML) resolve problem; a) use virtual machines (VM) [Clark et al,2005] analysis technology and desktop VM clones may help to resolve problem; b) automated QA failure protocols and selection options of user-delivered algorithms is one of the possible ways to resolve problem; c) and finally orientation of man-machine interfaces and overall “OpenPhotogrammetry” design on academic and research (rather than production) use will help to resolve, at least partially, problem d). Learning games elements will also be considered. Another problem which has to be resolved is integration of open-source image compression algorithms for the large imaging files transmission. Medical informatics research [L. Rodney Long], [B.Read] demonstrates some open-source solutions which can be deployed in geospatial domain. The following paper sections outline foreseeable solutions to the stated above problems.
3. OPEN-SOURSE SOFTWARE AS “OPENPHOTOGRAMMETRY” SERVICE
Recent years demonstrated great potential of social networks as efficient collaborative environments capable to even to catalyze revolutions such as “Arab spring”. Samples of efficiently functioning collaborative environments include: NSF Earth Cube[NEC], NASA Earth Exchange[NEX], NASA World Wind[WorldWind], Google Earth[ GE] and other virtual globes [Elvidge and Tuttle 2008]. Facebook, Google+, DropBox and ICloud become a standard way of communication and data exchange worldwide. Specifically LinkedIn [LIN] professional social network already has 220 geospatial and 145 remote sensing related working groups. For being compliant with such ubiqiotous collaborative networks and cyberinfrastructures
”OpenPhotogrammetry” from system architecture standpoint should be implemented as service bus or as it is depicted on Fig. 5. ”Openphotogrammetry” system architecture comprises the following components:
- Sensor Modeling - to deploy and append interoperable sensor models and georeferencing of raw geospatial data;
- Learning/Teaching - for teaching and orthorectifying imagery with plug-ins to OpenGIS and collaborative virtual globes;
- Feature Extraction - for overlapped geospatial imaging orientation, 3D feature collection and DEM generation;
- OpenAT - for densification of the ground control by analytical triangulation.
“OpenPhotogrammetry” is foreseen as an open-source development within multi-tier client-server and standalone software architectures. Any photogrammetry or geospatial imaging specialist/researcher in the world will be able to contribute new sensor-models and custom processing algorithms for “OpenPhotogrammetry”. Finally “OpenPhotogrammetry” will be seamlessly integrated as a sub-framework into ubiquitous network collaboration tool with capability to execute “OpenPhotogrammetry” related activities even from smart phones.
Figure 5. “OpenPhotogrammetry” system architecture
4. SOME ASPECTS OF “OPENPHOTOGRAMMETRY” SYSTEM
ADMINISTRATING
To promote integration and consistency, our “OpenPhotogrammetry” architecture will inhibit open-source software plugins that we develop to coincide the application programming interfaces (APIs) of image repositories such as GoogleEarth, NASA Earth Exchange, NASA World Wind, etc. We will develop and enforce security in a standardized manner to control access of images (e.g. public versus confidential), along with the ability to log or track application's usage.
Plugin Development:
We plan to develop a Java image processing program for “OpenPhotogrammetry” will run either online as an applet or can also be downloaded. It will support all the major photogrammetric image types, such as Tiff,JPG,MrSID and will run on mainstream operating systems, such as Microsoft Windows, Mac OS, and Linux. It will also support multithreading, so image file reading and processing can be completed in conjunction with other tasks. This developed plugins will make it possible to pre-processe geospatial imagery for being integrated into collaborative environments such as Google Earth(via KML/KMZ) and NASA Open Wind(via Java). Image zoom capabilities and transformations such as scaling, rotation, and flipping will also be attributes of this software. The “OpenPhotogrammetry” software will be created using an Java built editor and Java compiler and customized to each image repository. End-users will use their Java-enabled Internet browser, as a client; to connect to the “OpenPhotogrammetry” server and either download or run online the applet that we have developed. Once this has been completed, users will be able to conduct photogrammetric processing and collaborate over an open-source platform.
Security and Logging Enforcement
To maintain security and logging of who has access to images accessed on the “OPENPHOTOGRAMMETRY” system, we plan to integrate the XACML [OASIS-XACML] standard within the proposed architecture. XACML is a security standard for XML data and will serve as a flexible and sound security solution within our open-sourced environment.
XACML Architecture
The “OpenPhotogrammetry” XACML architecture will consists of the following entities:
• PEP - Policy Enforcement Point
◦ Responsible for enforcing policies based on authentication of users and their respective authorization to access information/services. The PEP describes who/what/when/where is being accessed to the PDP for a confirm/deny response.
• PDP - Policy Decision Point
◦ Responsible for making admission control and policy decisions in response to the request from the PEP in regards to the policy store.
• PIP - Policy Information Point
◦ A storage point of the policies that will be accessed by the PDP when making a decision.
Decision Mechanism
The XACML PDP has three main contexts. The three are policies, requests and responses. All three of these are classified in XML format.
The rule evaluations used within the PDP is the most critical part of the “OpenPhotogrammetry”„s XACML architecture. This request will be matched against policies in place. This request consists of attributes which describe the event to be evaluated, Subject, Resource, Action, (optionally) environment. For example, if a user should have access view and make changed to confidential image, policies created regarding this user's attributes (e.g. name, rank, employment, status, etc.) would acknowledge these attributes and make a decision based on matching user's credentials and attributes to those associated with the image or metadata. The response from such a request will have a variety of values as seen in Fig. 6. We have classified the response to three values. The first two are self-explanatory, while the third classifies requests that do not correspond to the policy. This is useful for testing purposes and design.
Target Condition Rule Value
■"Match" "True'’ Effect
■"Match" ■False-’ "■NotApplicable"
■"Match" 'Indelerminale" ""Indeterminate"
■"No-mateh" Don't care ""NotApplicable"
“Indeterminate’ Don't care ""Indeterminate"
Table 4 - Rule truth table
Figure 6. Rule table
SAML
Security Assertion Markup Language (SAML) [OASIS-SAML] is an XML standard for exchanging authentication and authorization data between entities. It consists of a precise syntax and processing semantics in the form of assertions, usually based on categories such as role-based access control (RBAC), which contain statements that are used to make access control decisions. SAML assertions and protocol messages are encoded in XML and are embedded in other structures for transport, such as HTTP POST request or XML-encoded SOAP messages. End-users will be assigned SAML attributes based on their affiliation and rank to verify their access to confidential information.
Application Flow
When the application is launched, the end user will be directed to a web page that we have developed and requesting their login credentials. The user will then enter their username and password which and verified with the backend of the software. Based on their attributes, they will be only able to access to what they are permitted. The current system has two mains sections, the user interface which consists of the website and the back end of the PDP. The front end and back end the use were chosen carefully and with consideration.
The front end of choice was PHP. This was because we wanted to be able to produce a dynamic web site. PHP provides for many security features including built in functions for one-way hashing and other encryption algorithms. PHP also includes a system for transparent cookie encryption. All these features and functions are part of essentially setting up and maintaining a secure web environment.
The back end will be Java-based. Java was chosen because of its platform independence and currently available frameworks. This framework was implemented to allow for the correspondence between PHP and a PDP. The user interface is used to interface with the user and the “OpenPhotogrammetry” image and metadata repository. It can read and write request XML files. This allows for dynamic on-the-
fly requests to be formed and be prepared to be sent to the PDP. This system uses a
system call to interface with the Java backend. Then the XML response returned from the backend will be processed.
Logging
There are many areas throughout the application that actions are logged into the database for auditing. This is to provide a necessary means of keeping track of access to the system and the information and images that reside on the “OpenPhotogrammetry”'s design. The following actions are currently planned to be logged:
- Logging in
- Incorrect Password
- Meta-data access
- Image access
- User Access
CONCLUSION
Success in “OpenPhotogrammetry” development will beneficial for researchers of ISPRS, FIG and many other research communities. Multi-year “OpenPhotogrammetry” research and development effort has potential of being forward-looking, integrative and transformative. “OpenPhotogrammetry” opens opportunity for fruitful international collaborations.
References
1. [Arpenteur] ARchitectural Photogrammetric Network Tools for Education and Research: http://sudek.esil.univmed.fr/arpenteur/spip/spip.php?rubrique7, accessed February 2012
2. [Botts 2005 ] Botts, M. OpenGIS® Sensor Model Language (SensorML). OGC 05-086. Open Geospatial Consortium Inc. http://portal.opengeospatial.org/files/?artifact id=12606
3. [BIAS] Basic Image Algorithms library: http://mip.informatik.uni-
kiel.de/~wwwadmin/Software/Doc/BIAS/html/index.html , accessed February 2012
4. [CATCON] Computer Assisted Teaching CONtest,
http://www.isprs.org/education/software.aspx , accessed February 2012
5. [Chen et al, 2011]N. Chen, Z. Chen, L. Di, and J. Gong, “An efficient method for near realtime on-demand retrieval of remote sensing observations,” IEEE J. Sel. Topics Appl. Earth Observ. Remote Sens. (JSTARS), vol. 4, no.3, pp. 615-625, Sep. 2011.
6. [Clark et al,2005] C. Clark, K. Fraser, S. Hand, J. G. Hansen, E. Jul,C. Limpach, I. Pratt, and A. War eld. Live Migration of Virtual Machines. In Proceedings of the 2nd ACM/USENIX Symposium on Networked Systems Design and Implementation (NSDI), Boston, MA, May 2005.
7. [Deng and Di, 2010] Deng, M., and L. Di, Facilitating Data-intensive Research and Education in Earth Science - A Geospatial Web Service Approach. LAP LAMBERT Academic Publishing GmbH & Co. KG, Saarbrbcken, Germany. ISBN: 978-3-8383-9714-6.
8. [Di et al 2003] Di, K., R. Ma, and R. Li, 2003. Rational functions and potential for rigorous sensor model recovery, Photogrammetric Engineering & Remote Sensing, 69(1): pp. 3341.
9. [Di 2005] L. Di “A Framework for Developing Web-Service-Based Intelligent Geospatial Knowledge Systems,” Journal of Geographic Information Sciences. Vol. 11, no 1, pp. 24-28.
10. [Di 2004] L. Di, GeoBrain-A Web Services based Geospatial Knowledge Building System. Proceedings of NASA Earth Science Technology Conference 2004. June 22-24, 2004. Palo Alto, CA, USA (8 pages. CD-ROM).
11. [Elvidge and Tuttle, 2008] , C D. Elvidge, B.T. Tuttle “HOW VIRTUAL GLOBES ARE REVOLUTIONIZING EARTH OBSERVATION DATA ACCESS AND INTEGRATION”, The International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences. Vol. XXXVII. Part B6a. pp 137-140, Beijing 2008
12. [GBM] Google Building Maker - http://sketchup.google.com/3dwh/buildingmaker.html, accessed February 2012
13. [GRASS] GRASS GIS (Geographic Resources Analysis Support System): http://grass.fbk.eu/, accessed February 2012
14. [Grodecki, 2001] Grodecki, J., 2001. Ikonos stereo feature extraction - RPC approach. Proceedings of the ASPRS Annual Conference, St. Louis, 23-27 April, unpaginated CD-ROM.
15. [Insight3D] Free photogrammetric 3D modeling open-source software: http://insight3d.sourceforge.net/, last accessed January 2012
16. [LDP] Learning Digital Photogrammetry web-site: http://ldipinter2.plan.aau.dk/, accessed January 2012.
17. [Levin et al, 2010] E. Levin, R. Liimakka, A. Leick, “Implementing a New Geospatial Educational Paradigm at Michigan Tech” Surveying and Land Information Systems, Vol.70, No 4, pp 211-216, Dec 2010
18. [Lillesand et al, 2000] Lillesand T., Olsen T., Gage J., McEnany P., ”New paradigm, new approaches: restructuring geo-spatial information education and training in a traditional research university settings.” International Archives of Photogrammetry and Remote Sensing, Vol. XXXIII, Part B6, Amsterdam 2000.
19. [Longley, 2002] LONGLEY, P.A., 2002, Geographical information systems: Will developments in urban remote sensing and GIS lead to ‘better’ urban geography? Progress in Human Geography, 26, pp. 231-239.
20. [B. Read] Berkeley Researchers Develop Technology for Sending Medical Images Via Cellphones http://chronicle.com/blogs/wiredcampus/berkeley-researchers-develop-technology-for-
sending-medical-images-via-cellphones/3891 , accessed March 2012
21. [L. Rodney Long] Transmission of Medical Images over Wide Area Networks: http://www.nlm.nih.gov/research/visible/vhp conf/long/vh96.htm , accessed March 2012
22. [LIN] http://www.linkedin.com/groupsDirectory?results=&sik=1330223055536
23. [NASA OpenSource ] http://ti.arc.nasa.gov/opensource/, accessed March 2012
24. [NEC] http://www.nsf.gov/geo/earthcube/ , accessed March 2012
25. [NEX] https://c3.nasa.gov/nex/ , accessed March 2012
26. [OASIS-XACML] OASIS Standard. “Extensible Access Control Markup Language (XACML), Version 2.0”. United States. 1 Feb 2005
27. [OASIS-SAML] OASIS Standard. “Assertions and Protocols for the OASIS Security Assertion Markup Language (SAML), V2.0”. United States. 15 Mar 2005
28. [OGC] http://www.opengeospatial.org/, accessed March 2012
29. [OpenDragon] OpenDragon project web-page http://www.open-dragon.org/open object.html, accessed February 2012
30. [PHT] Python Photogrammetry Toolbox http://www.arc-
team.com/wiki/doku.php?id=open software ppt-gui, accessed January 2012
31. [Petrie] Dr. Gordon Petrie Web-Links Database available online at: http://web2.ges.gla.ac.uk/~gpetrie/linksdb.html, accessed February 2012
32. [SPE] StereoPhotogrammetry Extractor : https://l aunchpad.net/ stereophoto
33. [Toutin, 2000] Thierry Toutin and Philip Cheng, “Demystification of IKONOS,” GIM Magazine, no 3, pp. 3-12, 2000
34. [WorldWind] http: // goworl dwind.org/features/
© E. Levin, G. Hembroff, 2012
