A simple Web browser provides the interface to the OMAR system. The map can be panned and zoomed to identify the area of interest. Date, time and other fields can be used to filter the desired search results.
Tabbed pages of metadata and an overview thumbnail are returned from the user’s search request.
Views are calculated and presented on the fly with full orthorectification and precision terrain correction. Data courtesy of DigitalGlobe Corporation.
The inset shows a resampled area of interest without image enhancements. Simple sharpening and histogram stretch operations can be applied by the user to bring out detail. Linear and area measuring, view export to file, band selection, and output to various streaming protocols are all supported to enhance the user’s workflow. Data courtesy of DigitalGlobe Corporation.
At the GeoInt Symposium in 2010, National Geospatial- Intelligence Agency (NGA) Director Letitia Long provided a vision for the future and labeled it GeoInt 3.0. That vision includes remote user access to imagery and products, collaboration, and manipulation of rapidly evolving agile technologies. Omar, from RadiantBlue Technologies, is a web-based processing and distribution solution that addresses many of the GeoInt 3.0 goals.
The solution embodies many of the features and approaches that were eloquently communicated by Director Long. She stated, “I want to put the power of GeoInt directly in the hands of our users.”
Today, Omar is deployed on classified networks providing remote access of imagery and video to the warfighter. Ships at sea, rescue missions, numerous patrols, and regular service personnel are routinely reporting how these quick looks have assisted their missions on a daily basis.
Editor’s Note: See also related article from The HumanGeo Group by Abe Usher.
The way in which this solution was developed and continues to evolve rapidly is just as interesting as its functionality. OMAR is open source software. It is developed and maintained in an online, distributed, and unclassified environment. The only exceptions are modular classified “plug-ins” that are maintained separately in classified networks. This architecture and approach allows development and integration at “Internet speed.”
The rapid evolution of this system is made possible through the use of open source software projects, policies, and procedures led by the Open Source Geospatial (OSGeo) Foundation. OSGeo is the leading organization for open source software projects. OSGeo provides governance, standards, practices, code review, automated testing, and configuration management for leading open source geospatial projects. These services are managed through a stringent incubation process, project steering committees, and community peer pressure. Most projects are commercially backed by companies and consultants. The resulting systems are more efficient, with lower life-cycle costs; just as important, they are extremely agile. There are now hundreds (if not thousands) of companies dependent on OSGeo technologies for the services and value-added products these companies bring to market.
This approach to providing services is in stark contrast to traditional government software procurement and development practices. Existing government procurement practices evolved from large weapons systems and are not appropriate for rapidly evolving commodity technologies. Historically, government acquisition led to pre-defined requirements decomposed into tasks and sub-tasks bound by rigid schedules. Rapid innovation is difficult in that environment.
Modern day open source practices and tools provide automated tracking of discrepancies and enhancements, self-documenting code structures, unit testing, continuous builds, and extremely rapid evolution. The Internet and most of the corporate world now run on open source software and have adopted these practices. The need for rapid technical evolution and the pressing need to reduce operating budgets have focused government attention on the open source approach.
Several government agencies are rapidly adopting open source software solutions and practices. As an example, NGA InnoVision (the NGA division tasked with implementing new technologies) recently announced a major initiative for open source software adoption. The primary drivers are total cost of ownership and technical agility. Recently we have seen the open source model applied to hardware design, such as the DARPA vehicleforge.
Companies such as RadiantBlue Technologies provide professional services and support for the resulting systems and bridge the requirements of government acquisitions to the commercial practices, domain expertise, and support of open source business models. These new business models earn revenue by providing value to the customer and flexibility through rapid sub-contracting with other open source companies and consultants.
Open standards and interfaces have allowed these systems to continually integrate and evolve to meet the operational challenges of constantly shifting missions. Significantly, these approaches allow the data to be manipulated and viewed remotely — often avoiding bandwidth limitations and data duplication that are prominent with “push” architectures. The Internet is constantly showing us new business models and ways to collaborate; many of these approaches have been applied in the Omar system.
Key features of the OMAR system are:
Remote discovery, viewing, and manipulation of imagery and products
On-the-fly orthorectification, precision terrain correction, and sensor model projection
Full Motion Video discovery and playback
ITAR-approved core baseline
Operation in both classified and unclassified systems
Standards-based interfaces to external systems and tools
No software installation needed at the user end
No licensing fees for the open source software system
The user needs only a web browser and network access to use the system. The design strategy has been to keep the interface simple with only a web browser required on the user’s device. All of the advanced processing and storage are managed on the server end where the data reside. This approach works well over long distances with limited bandwidth, providing a simple interface on the user’s end and concentrating all of the advanced technology, storage and processing upstream.
After logging in to the system, the remote user can interactively pan and roam through a reference map of the world to the area of interest. Footprints of available datasets are shown on the map, and results can be filtered based on a wide range of metadata parameters including date, time, sensor, target ID or any arbitrary combination of metadata values.
The user needs only a web browser and network access to use the system . . . All of the advanced processing and storage are managed on the server end where the data reside. This approach works well over long distances with limited bandwidth, providing a simple interface on the user’s end and concentrating all of the advanced technology, storage and processing upstream.
When the user presses the search button, a results page is returned. The results page includes all of the available datasets that meet the user’s criteria. See Figure 1.
An overview thumbnail is included along with a listing of the associated metadata. Tabs can select additional information about the dataset and options for streaming the data. Clicking the thumbnail allows the user to roam and zoom into the image or product at will. Additional controls are available for image adjustment. The user’s cursor displays the location readout in a number of formats. See Figure 2.
Many users of the system are using these interfaces to stream OMAR data into Google Earth. The system includes downloadable network links that automatically will fetch the latest data as Google Earth users change their view.
Simple controls for brightness, contrast, and sharpening can be adjusted. The resulting view can be saved locally in a number of formats for use in briefings or as input to other tools. Additionally a number of standard web mapping services allow OMAR views to be fed directly into other applications. Remote viewing and manipulation of National Technical Means (NTM), NGA products, Commercial Imagery, and Motion Imagery Standards Board (MISB)-compliant Full Motion Video (FMV) is provided by the system. NTM imagery is orthorectified and precision terrain corrected on the fly and projected into the web page view with the addition of a software plug-in that is separately maintained in a classified environment.
Once the image is selected, the user can roam and pan at will through the full size remote image. See Figure 3.
Most users are simply looking for reference or recent information in a very small area of interest. The system has exposed simple controls for manipulating the view, reading the coordinates, and taking simple measurements. The user can also save that adjusted view locally in a number of geospatial file formats. The goal was to put the users in charge and give them direct access to their area of interest in the national archives. See Figure 4.
The OMAR system has been deployed on distributed super-computer networks, simple servers, and personal computers. The general concept is to move the software to the data, not the data to the software. Continuing evolution of social networks, standards-based protocols, and online technologies will be integrated into new releases of the system.
In summary, Omar is a demonstration of a new approach of software development and user interaction for everyday government users. The open source software development model and rapid innovation on the Internet lead to technical agility and dramatically lower total cost of ownership for remote manipulation and viewing of remotely sensed data.