Image Access Solutions (IAS)
A Platform for Image Processing and Dissemination
Introduction
As applications for geospatial imagery have grown, so too has the need to efficiently manage and deliver imagery across an organization. This need is particularly important in enterprise environments, where organizations need to provide these key imagery capabilities across an increasingly distributed and heterogeneous network of users.
The capabilities of earth observation sensors continue to evolve. Sensors such as WorldView-1 and GeoEye provide data with higher resolution and more bands resulting in larger file sizes. This presents a problem to users requiring immediate access to the data when confronted with either low bandwidth communication links or network contention issues due to a large volume of users.
Exelis VIS has a solution in Image Access Solutions (IAS), which provides a comprehensive toolset for fast compression, access, and delivery of image files. These large imagery datasets, after being compressed to JPEG 2000 format, can be accessed via JPIP (JPEG 2000 Interactive Protocol). The combination of this compression format and protocol allows users to access specific regions of an image at a specific resolution without downloading the entire dataset. This means that users in bandwidth-constrained environments can have access to high- quality imagery over the Internet.
This paper discusses the Enterprise IAS product suite. It should be noted that understanding Enterprise IAS may not only require some prior domain knowledge of imagery standards and formats, but also an understanding of the assortment of technologies that breathe life into an enterprise-level solution. This document provides some background information by way of introduction, but the reader is encouraged to follow the provided references when a more thorough explanation may be required
History of IAS
This section will introduce the reader to the core concepts behind IAS and the evolution of the product from proof-of-concept to the enterprise level product suite that is Enterprise IAS.
What is IAS?
IAS is a product suite consisting of both client and server components for high-performance image storage, compression, conversion, and delivery based on the JPEG 2000 standard. JPEG 2000 is a compression format based on wavelet technology that outperforms older standards like JPEG in compression efficiency and functionality. VIS has more than 20 years experience in image compression and has contributed significantly to the development of the JPEG 2000 standard.
JPEG 2000 differs from other compression standards because it embeds various quality levels and resolution levels within a single compressed bit stream in a single stored file. From the user’s perspective, this means that any resolution up to the original resolution can be extracted from the compressed file by simply pulling the necessary bits from the wavelet bit stream. Multi-gigabyte images can be viewed on demand, regardless of bandwidth or latency constraints or available RAM on the client device.
On the server, IAS provides image compression, storage, conversion, and delivery. On the client side, IAS provides a high-performance, user-friendly interface for viewing and chipping the streamed imagery.
IAS Evolution
IAS was initially developed as research code jointly by ITT Geospatial Systems division (ITT GS) and Visual Information Solutions (VIS) based on Kakadu (http://www.kakadusoftware.com), an implementation of the JPEG 2000 standard developed out of the University of New South Wales in Australia. VIS continued to add capabilities to IAS until it was a fully-featured Commercial Off-The-Shelf (COTS) product. The first commercially available versions featured:
- NITF file support
- Department of Defense Intelligence Information System (DoDIIS)/Joint Interoperability Test Command (JITC) compliant
- Support for large image files. Images that are considered large varies widely based on the user, but can generally be defined as images larger than 1 MB in size.
- Authentication
- Clustering and load balancing
- An Administrator Console
- Java interface to compression engine and server farm
At IAS version 3.0, it became clear that when integrating with existing enterprise architectures, IAS presented several obstacles. Authentication, clustering, load-balancing, and administration were all non-standard, proprietary implementations, and the JPIP streamer was implemented as an embedded custom web server, meaning that the server handled all HTTP requests instead of employing already available web servers such as Apache. The embedded custom web server within IAS 3.0 implemented a subset of the HTTP 1.1 specification and, as such, limited the use of proxy servers and secure streaming (HTTPS). Furthermore, this stovepipe architecture made it nearly impossible to leverage existing server capabilities (such as authentication, clustering, and load-balancing) in an enterprise environment.
Due to these limitations, VIS made significant architecture changes and renamed the product line to Enterprise IAS (E-IAS), the current IAS Platform.
Enterprise IAS
As more organizations move to service oriented architectures, the IAS product was developed into an enterprise component to streamline integration into enterprise environments. Enterprise IAS (E- IAS) is a Java Platform Enterprise Edition (Java EE) component. As such, E-IAS can leverage existing server technologies such as security, load balancing, and web serving, which in previous versions of IAS had been proprietary, closed implementations. Additionally, the following improvements were made:
- Removed the custom web server for JPIP and implemented the JPIP protocol within a standard web container
- Wrapped ENVI in a Java message-driven bean for JPEG 2000 compression
- Added transcoding (see Section 4.3), TFRD to NITF conversion, Web Coverage and Web Mapping Services (WCS and WMS)
- Added a simple HTTP interface for holdings management and discovery
Benefits of EJAS
Many of the benefits of Enterprise IAS can be attributed to its standards-based approach and compliance with Java EE standards.
- Secure Streaming - HTTPS/SSL, also known as JPIPS
- Standard administration - Authentication, clustering, load balancing, failover, bandwidth control, proxying, and user administration are all handled by the controller or application server already in use.
- Multiple stream-time tap points - Enterprise IAS provides interceptors for on-the-fly operations such as metadata insertion and filters for limiting connections and supporting releasability.
- Metadata Insertion - On-the-fly metadata insertion means greater flexibility in providing users with timely and highly granular metadata control for streamed imagery. For example, different metadata could be provided based on the type of classification of user.
- Multiple platform and application support - Because E-IAS is a Java EE component, integration into your enterprise is made easier. E- IAS already supports JBoss, WebLogic, and Glassfish, but it could also support any other Java EE compliant application servers.
- NPJE (NSIF Preferred JPEG 2000 Encoding) to EPJE (Exploitation Preferred JPEG 2000 Encoding) transcoding - ENVI performs transcoding in the wavelet domain, eliminating the need to uncompress before conversion between formats.
- REST API - This is a standard API that allows users to leverage freely available 3rd party client libraries or to create their own libraries to administer, use, and manage E-IAS.
The benefits outlined above not only directly impact ease of integration, performance, and configurability, but also translate to a lower cost of ownership. Specifically, the following characteristics can have a significant impact on costs associated with the deployment, administration, and use of the system:
- Ability to optimize communication bandwidth using robust metadata control
- Ability to increase the performance to price ratio using Linux
- Lower integration costs by using standard technologies well-known to system administration staff
Enterprise IAS in Detail
The Enterprise IAS suite consists of the E-IAS server and complementary client components. Client components include a light-weight viewer and the IAS Client Chipper, a tool for saving a JPIP stream to NITF, allowing the user to save a subset of an image to a local file. Client components are considered optional. For example, any viewer that can accept a JPIP stream can be used with the E- IAS server. The Chipper would only be required by those who require uncompressed subsets of images on their local machine.
The server consists of four core components: the JPIP streamer, the compressor, the transcoder, and the registry. Figure 1 illustrates the relationship of the IAS components. Each component is described in detail below.
E-IAS Components
IAS JPIP Server
The IAS JPIP Server enables an application server to stream imagery via JPIP over http or https. Additionally the JPIP server supports on-the-fly insertion of JPEG 2000 metadata boxes (e.g. geo-reference parameters) into a client-requested JPIP stream. The IAS JPIP Streamer is implemented as a stateful Enterprise Java Bean. Secure streaming is enabled by configuring SSL on the application server or on a front-end controller, such as Apache. Supported image types for streaming are JPEG 2000 (including GeoJP2 or GMLJP2), and NITF 2.1 C8.
Metadata Insertion
JPEG 2000 metadata boxes can be inserted on-the-fly into a client-requested JPIP stream. Metadata insertion is decoupled from the JPIP engine allowing for highly granular control over image metadata. This means that metadata can be updated at any time, or different sets of metadata can be streamed for different users. Also, different products can then be supported. For example, geo- referencing information can be encoded as either GMLJP2 or GeoJP2.
The metadata resources can be local files, URLs to other metadata (e.g. GeoRSS metadata insertion from a temporal weather feed) or data that is kept in the database. The IAS Registry manages the associations of metadata resources with datasets and vice versa.
WCS and WMS Services
Enterprise IAS also provides standard OGC WCS and WMS interfaces to retrieve imagery for clients that are not JPIP enabled. Supported outputs through these services include NITF, GeoTIFF, JPEG, and PNG. If NITF is requested, it will always be NITF C8.
WCS
- Supports version 1.0.0 requests
- Client can specify which bands are to be included in the returned coverage
WMS
- Supports version 1.1.1 requests
- Assumes that images are already orthorectified
- Only a single map layer (the image) is returned
- Returns an image with 8-bits per channel(For a monochromatic image this results in 8-bit space independent of the radiometric resolution of the sampling sensor. For color images, the sampling space is 3 times 8 bit with 8 bits for each red, green, and blue band)
IAS Server Registry
The IAS Registry is a registry built on a simple data model based on the Atom publishing protocol for registering datasets and Atom feeds for disseminating available data, metadata and associations via HTTP.
The IAS Registry is a simple registry based on the Atom publishing protocol for registering datasets, submitting compression jobs, and providing Atom feeds for disseminating available data, metadata and associations. The registry is a stateless transactional component and replaces interacting with the server database through the database SQL interface. Adding data using database scripts is still possible; the registry provides a simpler interface and is dependent only on the availability of the http protocol within the development tool of choice.
Atom Feeds / RSS allow the user to query the server for information via http(s), only a browser is needed. To publish data to the database, such as registering an image or image metadata, one would need to post the appropriate atom document.
The IAS Registry currently supports:
- Atom publishing of datasets and metadata
- Paging of dataset and metadata collections
- Associating metadata with data and vice versa
- Metadata of category type POLYGON EPSG:4326 WKT (Well Known Text) will result in a GeoRSS / GeoAtom feed in the registry browser. For example:
POLYGON((-78.93 33.68, -78.93 33.68, -78.93 33.68, -78.93 33.68, -78.93 33.68))
- Metadata of category type SUMMARY will result in metadata being displayed in the registry browser
- Metadata of category type XML will result in the first associated metadata entry for a particular dataset being inserted in the requested JPIP stream.
- Compression job submission and monitoring
- OpenSearch (http://www.opensearch.org) - standard way of searching the Registry's holdings.
E-IAS Compressor
The IAS Compressor is an asynchronous scalable component that will compress NITF, TIFF, and GeoTIFF data to NITF 2.1 with JPEG 2000 compression (NITF C8). Compression output can be controlled through the use of compression profiles to meet a variety of compression needs.
The IAS Compressor will compress NITF 2.0/2.1, TIFF, and GeoTIFF images to the NITF 2.1 C8. The IAS Compressor is implemented as a Java Message Service (JMS) message driven bean (see Section 3.2.2). The IAS Compressor is invoked when a compression job is submitted to the IAS Registry via the Atom interface. The JMS service of the application container then manages the life cycle of each compression job by assessing the priority of each job. It can even be configured to re- try jobs in the event a compression job fails to begin due to unavailability of resources, or exits before completion due to an error in the execution of the job.
The way an image is compressed is determined by the compression profile used. Several profiles are included with an Enterprise IAS distribution. Compression profiles must be registered in the IAS Registry via the Atom interface so that they are available to the IAS Compressor. IAS provides several pre-configured profiles, which are described in the following section.
Compression Profiles
The IAS Compressor includes pre-configured compression profiles suitable to meet a variety of compression needs. The profiles are:
- Huge image compression without tiling
- Large image compression without tiling
- Lossless image compression
- Lossy image compression
- Lossy image compression with a target rate of 2.4 bits-per-pixel
- PAN image NGA Exploitation Preferred JPEG 2000 Encoding numerically lossless compression
- PAN image NGA Exploitation Preferred JPEG 2000 Encoding visually lossless compression
- SAR image NGA Exploitation Preferred JPEG 2000 Encoding numerically lossless compression
- SAR image NGA Exploitation Preferred JPEG 2000 Encoding visually lossless compression
- PAN image NGA NSIF/BIIF/NITFS Preferred JPEG 2000 Encoding numerically lossless compression
- PAN image NGA NSIF/BIIF/NITFS Preferred JPEG 2000 Encoding visually lossless compression
- SAR image NGA NSIF/BIIF/NITFS Preferred JPEG 2000 Encoding numerically lossless compression
- SAR image NGA NSIF/BIIF/NITFS Preferred JPEG 2000 Encoding visually lossless compression
- Tiled image compression using 1024 x 1024 tiles
E-IAS Transcoder
The IAS Transcoder converts images between the Exploitation Preferred JPEG 2000 Encoding (EPJE) and the NITF Preferred JPEG 2000 Encoding (NPJE) format, without having to uncompress and re-compress the image. Often NPJE is the preferred encoding format used when collecting data. However, EPJE is the preferred format for streaming as it facilitates the rapid access to a variety of resolution levels. It is also the preferred encoding for use with many COTS analysis products. The IAS Transcoder transcodes between these two formats in one convenient step, saving the user from de-compressing and re-compressing the image to reach the desired format. Compression and Transcoding duties are both performed by an engine based on ENVI.
ENVI
ENVI is a commercial-off-the-shelf (COTS) software solution for processing and analyzing geospatial imagery. While ENVI is commonly used within a desktop application, many of its underlying components and algorithms can be wrapped such that they can be deployed in an enterprise architecture as stand-alone services. For example, the default compression engine that ships with Enterprise IAS is based on ENVI. Enterprise IAS also uses ENVI functionality for NITF file creation.
Additional ENVI functionality, such as orthorectification, mosaicking, pan sharpening, change detection, and other imagery and geospatial data processing tools can also be wrapped for deployment as services in the enterprise environment. Recently, ENVI based services for anomaly detection, transcoding, and image chipping have been developed and deployed as add-on capabilities. These services are not included in the baseline Enterprise IAS offering but can be purchased as plug-ins to Enterprise IAS. Additional ENVI based services may be included in future versions, or they can be added as custom services depending on the customer’s needs.
E-IAS Client Components
There are a number of client-side options available for accessing data provided by the E-IAS server. VIS provides its own client-side components, but there are also open source SDKs that can be used to create a customized viewer to meet more specific needs.
Enterprise Viewer
The Enterprise Viewer is a freely available and distributable application that allows a user to view JPIP streams and includes some additional functionality to optimize image viewing. VIS has a freely available JPIP viewer, the Enterprise Viewer, a standalone Java application which can view any standard JPIP stream. Among the Enterprise Viewer’s core features are:
- Zoom
- Image Rotation
- Dynamic Range Adjustment
- Image Sharpening
- Establish waypoints and Waypoint following
- Save current view as JPEG/GeoTIFF
- View codestream parameters
- North arrow annotation
- Tabbed interface for viewing multiple images
The Enterprise Viewer includes additional functionality when used with Enterprise IAS. These features are particular to the Enterprise IAS family:
- Security/releasability annotation for NITF imagery
- Save to NITF via the Enterprise IAS WCS service
The Enterprise Viewer can also be embedded into a web page as an applet or the entire application can be deployed via Java Web Start. Java Web Start is a technology included with the Java runtime environment that enables users to download client applications from a central location. If the application is updated, Java Web Start will detect the change and automatically update the client machine.
IAS Client Chipper
This tool will save a JPIP stream to uncompressed NITF or GeoTIFF onto the client’s machine. This component is typically invoked from the Enterprise Viewer. This purchasable component is invoked through the Enterprise Viewer and allows a user to chip a region of interest from the image directly from the JPIP stream to the client machine in uncompressed GeoTIFF, NITF 2.0, or NITF 2.1 format from a NITF C8 or GeoJP2 file. The image chip preserves the raw data of the original image (assuming lossless compression), including metadata (e.g. geo-reference parameters), which can then be used in other image processing tools.
Open Source Client SDKs
VIS has also widely contributed to provide JPIP client libraries for the open source community. These libraries are freely available for users to create JPIP clients to meet their specific requirements.
GDAL
The Geospatial Data Abstraction Library (GDAL, http://www.gdal.org) is a translator library for raster geospatial data formats. ITT VIS has released to the GDAL community a JPIP client format driver which integrates into the GDAL architecture delivery of image data and metadata from a plain and SSL enabled HTTP JPIP server. The JPIP client format driver interprets the metadata in both the GeoJP2 and GMLJP2 formats streamed from the server.
The GDAL JPIP format driver can be used in the GDAL supported programming languages, which include Java, Python, C#, Ruby, C++ and C. Currently there are tested GDAL JPIP format driver interfaces in Java, Python, C++ and C.
Java SDK - ImageIO
Java’s ImageIO API provides a pluggable architecture for working with images that allows for the addition of format-specific plug-ins. ITT VIS provides an open source JPIP ImageIO interface to GDAL and closed source ImageIO interfaces for Kakadu and pure Java JPIP client libraries. The advantage of this approach is being able to plug different ImageIO JPIP providers into existing java clients that support this interface definition. For more information on ImageIO visit: https://jai-imageio.dev.java.net/
Security
The Java Authentication and Authorization Service (JAAS) is a low-level API for adding flexible authentication and authorization services. Java EE compliant containers implement the JAAS specification. In addition to the JAAS provider supplied in the deployed application container, Enterprise IAS uses servlet filters to secure the main access points to the JPIP Streamer EJB by using Spring Security system (http://static.springframework.org/spring-security/site/).
Spring Security has the following advantages:
- Easy development
- Enterprise-wide single sign on using Yale’s Central Authentication Service
- JAAS integration
- LDAP integration
- Channel security (redirect to HTTPS as required)
- Java 5 annotation support
- Caching
Leveraging Standards and the Enterprise
Enterprise IAS is built upon published standards and follows best practices for an enterprise-level application.
JPEG 2000 and JPIP
The Joint Photographic Experts Group (JPEG) is an international body that participates in the development of imagery file format standards. The JPEG Discrete Cosine Transform (DCT) compression standard developed has been widely adopted but is not without its limitations. Different resolutions, qualities, or spatial regions cannot be extracted without decompressing the whole image. Also, artifacts are generated at high compression ratios. To address these issues, a new standard, JPEG 2000, was developed.
The JPEG 2000 standard uses Discrete Wavelet Transform (DWT) technology. An important advantage of the JPEG 2000 format is that it supports both lossy and lossless compression formats. Another advantage of the new standard is the scalability of the codestream. Scalability is supported in four dimensions: quality, resolution, region of interest, and band.
- Quality Scalability: A JPEG 2000 image can be encoded with varying quality layers. This means that two different users can view the same image at different quality levels. One analyst may require viewing the image at full quality for detailed analysis, while another only needs to request the image at lower quality level just to get the a rough impression of the image.
- Resolution Scalability: Fast access to multiple resolutions allows clients to quickly zoom in and out when viewing an image.
- Region of interest coding: Clients will receive only the specified region of interest requested which can be sent to the client without decoding the compressed image.
- Band: For multi-component images, clients can request to receive only the components of interest.
Part 9 of the JPEG 2000 specification defines the JPEG 2000 Interactive Protocol (JPIP). JPIP is a client-server protocol than can be implemented over HTTP and exploits the random access and scalability of the JPEG 2000 standard in a networked environment. Additionally, the data does not have to be decompressed before sending to the client.
The combination of JPEG 2000 and JPIP enables networked clients to have very fine control over the data that they receive. Clients have random access to different image regions and can zoom and pan the image without decoding the entire image. Because only a portion of the image is sent to the client, even bandwidth-constrained users are able to have access to high quality imagery.
For more information on the JPEG 2000 standard visit: http://www.jpeg.org/jpeg2000/index.html
Java Enterprise Edition (Java EE), Enterprise Java Beans (EJB) and the Java Message Service (JMS)
Java EE is the industry standard for developing portable, robust, scalable and secure server-side Java applications. Building on the solid foundation of Java Standard Edition (Java SE), Java EE provides web services, component model, management, and communications APIs that make it the industry standard for implementing enterprise-class service-oriented architecture (SOA) and next-generation web applications. For more information on Java EE visit: http://java.sun.com/javaee/.
An Enterprise Java Bean (EJB) is a server side component that encapsulates the business logic of an application (The Java EE 5 Tutorial: http://java.sun.com/javaee/5/docs/tutorial/doc/bnblt.html.) In E-IAS, EJBs encapsulate the JPIP session as well as the compression and transcoding services. This implementation allows server performance to be optimized since EJBs are designed to swap between memory and persistent storage to manage load as required. EJBs are easily managed by the Java EE container for easy administration of object pools and life cycles. Implementation is vendor neutral and scales to multi-core and multi-CPU machines.
The E-IAS compression and transcoding services are implemented as message driven beans, a specific type of EJB that can receive messages through the message manager, the Java Message Service (JMS). JMS is a service implemented by a Java EE compliant application server. JMS gives the message driven beans the ACID features of:
- Atomicity
- Consistency
- Isolation
- Durability
These are the key transaction properties of a reliable database system. In addition, JMS is asynchronous and supports the concepts of queues, clustering and provides a scalable architecture for the intensive task of handling multiple compression and transcoding jobs.
Representational State Transfer (REST) and Atom
REST is an architectural style that broadly describes how to define and address information or resources, where resources are uniquely addressable. A REST approach is referred to as a Resource Oriented Approach (ROA) and contributes to an Enterprise System. E-IAS takes a RESTful approach for managing data and metadata in the IAS Registry by following the Atom Publishing Protocol. Atom is a well defined format for managing web resources through basic operations supported by the HTTP protocol. For more information visit:
OpenGIS
The Open Geospatial Consortium (OGC) defines standards for geospatial and location based services. E-IAS implements both the Web Coverage Service (WCS) and Web Map Service (WMS) interfaces to provide alternative access to imagery over HTTP by non-JPIP clients. E-IAS defines the HTTP Key Value Pair interface to the WCS and WMS services and does not provide the SOAP interfaces that are also available through these standards. Within E-IAS the WCS and WMS interfaces provide a common level of interoperability with other systems.
WMS provides an interface for requesting geo-registered map imagery, typically returned in a format such as JPEG or PNG. Requests made through the WCS interface will return a geographical coverage, e.g. data encoded as NITF. For more information on these standards visit:
E-IAS in an Enterprise Framework
-IAS is flexible and can be deployed in a number of ways based on the system architecture and client needs. For some, a single instance will suffice. For those with a higher user load or data throughput, E-IAS can be clustered to meet the demand. Because of the number of variables involved, determining if an E-IAS cluster is needed is best left to be evaluated on a case by case basis. The following section covers what a typical E-IAS deployment might look like.
A typical enterprise IAS Deployment
In this scenario, E-IAS is deployed onto an existing JBoss application server cluster and integrated with an existing Oracle server that manages the image archive.
- JBoss, an open source Java EE compliant application server, is chosen as the deployment server for E-IAS. It is decided that a small cluster will be used to optimize performance and reliability.
- Apache, an open source HTTP server is selected as the front-end controller. Modules for load balancing and SSL for secure streaming have also been installed to improve reliability and security.
- E-IAS is deployed as a server component. Enterprise Java Beans for JPIP streaming, WCS, WMS, and access through E-IAS’s RESTful interface are managed by the application server.
- The security framework is configured to access authentication credentials through an existing LDAP directory.
- Integration with the existing image store is accomplished by harvesting image data into IAS tables that live in a dedicated Oracle database.
- Multiple clients and client types access Enterprise IAS. Some utilize Enterprise Viewer for viewing imagery via JPIP and JPIPS (the secure stream). Others may use a customized client based on GDAL or ImageIO. Finally, other custom clients perform tasks ranging from discovery to administration through the E-IAS REST interface.
Case Studies
IAS has been widely deployed and incorporated into a variety of commercial and government applications. Below, several real-world examples are described to illustrate IAS image compression and delivery services.
Providing Near-Real Time Images over the Arctic Ocean
Until recently, any journey in the ice-infested waters of the Arctic Ocean was a journey into the unknown, with no up-to-date, detailed map of the ice conditions readily available in such a remote region of the world. Now, Kongsberg Satellite Services (KSAT) provides Near Real Time (NRT) satellite imagery of the Arctic Ocean from the European Space Agency satellite, Envisat, and the Canadian Radarsat-1 satellite, to any International Polar Year (IPY) expedition in the Arctic Ocean.
Users will be able to connect to KSAT in Tromsø using an Iridium phone to access the latest radar satellite image of the sea ice conditions immediately ahead of them. The transmission of detailed satellite images over a very low network bandwidth is possible using IAS. IAS offers advanced compression and high performance delivery and management of geospatial imagery and data. Using IAS, coupled with KSAT’s ability to deliver the latest satellite image within half-an-hour of acquisition, users can receive the very latest information on the sea ice conditions.
Providing Tactical Users with Time-Critical Information
Today’s tactical forces need a way to access imagery residing in image archives often far from their operational location. Whether it is access to base maps or imagery that is only minutes old their missions are often dependent on imagery and feature data for situational awareness and making critical decisions. These users typically operate with a very bandwidth-constrained network. Traditional image delivery methods do not address the tactical timelines these forces require. The current paradigm requires the user to wait for the entire image file to be received before it can be opened for review or analysis. Additionally, transmitting entire geospatial images often hopelessly congest these already constrained networks, and in many cases the source imagery is too large for display on in-theatre devices.
IAS plays a key role in a system developed to provide tactical users the ability to quickly identify and retrieve high-resolution imagery over bandwidth as low as 5 kilobits per second (Kbps). The system is a complete solution for the rapid storage and dissemination of imagery data and is capable of disseminating exploitable data to users nearly 1000 times faster than conventional methods.
The system accomplishes this by leveraging IAS’s JPEG 2000 compression and JPIP streaming capability to quickly stream large format images across low bandwidth networks with zero loss of resolution, while allowing users the ability to interactively exploit an image during the transfer process. The increase in speed of dissemination to the analyst allows time critical mission decisions to be made.
IAS makes it possible for users of the system to quickly access a desired region of interest within a multi-resolution JPEG 2000 compressed image without having to receive or decompress the entire multi-gigabyte file. They can then view the region of interest at full resolution and quality by retrieving only 1–2 percent of a compressed file. This represents a download time of minutes as opposed to hours allowing the analyst to receive time critical information necessary to make mission decisions.
Commercial Data Provider Meets Customer Needs with IAS
A leading provider of high resolution commercial imagery must meet the needs of both commercial and government customers and provide each with prompt delivery of image products that meet their format requirements.
To better serve their customers, this provider integrated IAS into its production workflow. With IAS, the provider can support JPEG 2000 product delivery to commercial customers and NITF 2.1 C8 product delivery to government customers. They were also able to tailor their JPEG 2000 products for a specific customer through specific compression profiles supported by IAS.
Summary
As capabilities of earth observation sensors continue to evolve, so do the size of their products. Higher resolution and more bands are resulting in larger datasets of geospatial imagery. Providing timely access to this imagery across a base of heterogeneous users is an increasingly difficult problem, especially when the user may be in a bandwidth constrained environment.
VIS' product, Enterprise IAS, addresses this problem with a comprehensive toolset that provides compression, access and delivery of image files. Enterprise IAS is built upon the JPEG 2000 standard, a highly scalable compression format that offers both lossy and lossless compression. Enterprise IAS implements the JPIP standard to provide access to JPEG 2000 files in a networked environment. The JPEG 2000 format and JPIP protocol allow clients to access only the specified region of interest at the desired resolution, meaning that clients can view imagery without having to download the entire image.
Enterprise IAS is a Java Platform Enterprise Edition component. As such, E-IAS can leverage existing server technologies such as security, load balancing, and secure streaming. Integrating with existing standard technologies such as these could translate into overall lower cost of ownership, as they would not have to be re-implemented and should already be familiar to system administrators. This standards-based approach also allows E-IAS to be easily integrated with existing image stores. E-IAS leverages existing standards and the scalability of the JEPG 2000 format to allow organizations to deliver timely image-based information across a wide network, so that users can have access to the imagery they need to make informed decisions quickly.
Appendix
Enterprise IAS System and Support specifications
To ensure a successful deployment of Enterprise IAS it not only requires the right hardware and software, but also a knowledgeable support staff. At a minimum, Enterprise IAS can be successfully deployed and maintained by one individual with basic system administration skills on the target platform as well as prior experience with Java EE compliant application servers (such as those listed below). If the individual did not have prior knowledge of the various supporting technologies discussed in this study, it is strongly suggested that the person would spend some time to gain at least a basic understanding of them.
E-IAS currently supports the following configurations:
Abbreviations
References
For more information contact your representative at: info@exelisvis.com