December 2007

December 2007 | Executive compass

Looking forward to SOCET GXP v3.0

Mark Sarojak

It never ceases to amaze me how quickly the year can fly by, and it’s been another great year here at BAE Systems GXP. Among other notable events, this year marked the rollout of SOCET SET® v5.4 with its innovative Next-Generation Automatic Terrain Extraction (NGATE) functionality and the release of SOCET GXP® v2.3.1 with its transformational Spatially Enabled Exploitation (SEE) capability and live connections to both Google Earth and the ESRI® geodatabase. Equally as important, we have enjoyed the pleasure of welcoming many new customers to our BAE Systems GXP family in 2007. Yes, it has been an excellent year.

Now we are looking forward to 2008 and the achievement of a major milestone in the evolution of our products…SOCET GXP v3.0. As you may know, for the past several years we have been working hard to combine our image analysis software (VITec) and geospatial analysis software (SOCET SET) into a single product line. From the start, we understood clearly that the resulting product would have to have the accuracy and precision of SOCET SET while maintaining the ease and speed of VITec. Furthermore, with over 25 years of combined history and capability, we knew that integrating the two products wouldn’t be an easy task.

Despite all of the challenges, the goal is within reach. In mid 2008, SOCET GXP v3.0 will arrive with updated user interfaces, improved workflows, and greatly increased capabilities for geospatial analysis, as well as several key enhancements for image analysis. This upcoming release is exciting news for all of us, and I am sure that you are looking forward to it as much as I am. We will be certain to keep you updated as we get closer to the SOCET GXP v3.0 release, but feel free to contact us in the mean time if you have any questions.

SOCET GXP v3.0 combines image analysis and geospatial analysis capabilities in the same software package.

SOCET GXP v3.0 combines image analysis and geospatial analysis capabilities in the same software package

SOCET GXP v3.0 combines image analysis and geospatial analysis capabilities in the same software package

Also, as a reminder, registration is now open for our 2008 BAE Systems GXP International User Conference, April 7 – 11, 2008, at the Hilton La Jolla Torrey Pines in San Diego, CA. The conference is always a great opportunity for us to get together to share, learn, and build relationships, but this year’s conference promises to be even better with the introduction of a non-technical Executive Track and several new social events. Additionally, we will be showcasing SOCET GXP v3.0 with live demonstrations and workshops. I can’t stress enough how much we value your participation and feedback at our conference, and I look forward to seeing you all there. Just make sure to keep your eye on the calendar… April will be here before we know it.

Enjoy the holidays!!!

Sincerely,
Mark Sarojak
Mark Sarojak
Director of Sales, Americas
BAE Systems GXP

December 2007 | What's the buzz?

San Diego fires threaten Rancho Bernardo campus

A special thanks to all of our valued customers, industry partners, and colleagues for your concern and well wishes during the recent southern California fires this past October. All of us here in San Diego watched in horror as the flames, spread by extreme Santa Ana winds, threatened much of San Diego and nearby counties. We are grateful for the dedicated responders, relief agencies, and citizens who cooperated to limit the loss of life and property during these widespread fires. We are also proud of our company. BAE Systems announced a three-pronged approach to support employees, the relief efforts, and the first responders. This support illustrates an ongoing commitment to the local community, which is demonstrated year after year.

BAE Systems made a $100,000 donation to the American Red Cross’ Natural Disaster Relief Fund. The company also made a $50,000 donation to the California-based Fire Services Training Institute which was identified as the preferred beneficiary by the California State Firefighters’ Association. The Fire Services Training Institute is a California-based non-profit with the primary mission of bringing quality training to firefighters and promoting firefighter safety.

December 2007 | GXP in the news

BAE Systems provides intelligence software training to field analysts in Iraq and Afghanistan

Eric Bruce in a C-17 heading back to Qatar from Bagram Air Base.

Eric Bruce in a C-17 heading back to Qatar from Bagram Air Base.

BAE Systems GXP product specialists, Eric Bruce, Dennis Bryant, and Rob Stout, traveled to Iraq and Afghanistan recently to provide geospatial intelligence software training to field analysts.

The trip was part of BAE Systems formal partnership with the National Geospatial-Intelligence Agency’s (NGA) deployable systems upgrade team. In 2006, NGA purchased BAE Systems’ SOCET GXP software to support its customer, the Office of Global Support.

Rob Stout on board a C-130 in Afghanistan. Travel from site to site within Afghanistan required tactical capabilities of the C-130.

Rob Stout on board a C-130 in Afghanistan. Travel from site to site within Afghanistan required tactical capabilities of the C-130.

“The goal of the trip was to provide operational capability to use the tool,” said Dan London, BAE Systems vice president for Geospatial eXploitation Products. “At the same time, the information our team gathered in the field was invaluable, because seeing the product in-theater, and the issues analysts face, helps us refine the product.”

SOCET GXP is a geospatial intelligence tool that uses imagery from airborne and satellite sensors to identify ground features for improved situational awareness. Analysts in the field use maps and charts generated by SOCET GXP to perform before-and-after site comparisons and battle damage assessment, and to detect potential improvised explosive devices and ambush sites. The data can also be used to coordinate troop maneuvers, helicopter landings, and land-vehicle routes.

Dennis Bryant at former Saddam palace, Al Faw in Baghdad, now an operations center for U.S.commanders.

Dennis Bryant at former Saddam palace, Al Faw in Baghdad, now an operations center for U.S.commanders.

“The analysts were very attentive and interested in what we were talking about. Because they’re in an operational environment, they know once we leave, they are going to have to put that knowledge to use,” said Eric Bruce, one of BAE Systems product specialists deployed in the field.

As part of an overall software upgrade, SOCET GXP was installed for the first time on the rugged portable computers used by analysts in Iraq and Afghanistan.

The software processes data from a variety of image sources and creates image products that can be compressed, saved in multiple formats, and shared over secure networks. Data and reports can be immediately e-mailed and accessed from mobile laptop computers, relay stations, and ground control centers.

For the complete article and related webinar, Tales from the Warfront — Warfighters Receive Onsite Geospatial Training, GeoWorld Magazine, October, 2007, please visit:

December 2007 | SOCET GXP | Software update

SOCET GXP v2.3.1

Synchronized viewing with SOCET GXP and Google Earth

Synchronized viewing with SOCET GXP and Google Earth

SOCET GXP v2.3.1 features all of the functionality required to satisfy typical image analysis production workflows. This release adds seamless integration and synchronized viewing with Google Earth, and introduces a direct, bidirectional link to the ESRI geodatabase or SOCET SET feature database for dynamic viewing and editing of feature data. New in SOCET GXP v2.3.1 is Spatially Enabled Exploitation (SEE), designed to enhance image exploitation by creating attributed ground space graphics in a connected enterprise geodatabase environment. SEE allows the analyst to answer critical questions using spatial, attribute, and temporal queries. In addition, smart vector attribution supports external ESRI multi-user/personal databases, shapefiles, and SOCET SET feature databases. AutoSOCET adds an autonomous geospatial analysis workflow: automated triangulation, terrain data generation, orthorectification, and mosaicking.

SOCET GXP’s fundamental photogrammetric architecture is complete, and we are well on the way to full integration, moving the remainder of SOCET SET’s functionality into SOCET GXP, which has the same rigorous sensor models as SOCET SET, for highly accurate georeferencing.

December 2007 | SOCET SET | Software update

SOCET SET v5.4.1

ALOS

ALOS

SOCET SET v5.4.1 features a wealth of productivity enhancements for creating and editing high-resolution terrain and surface models, including improvements to the Next-Generation Automatic Terrain Extraction (NGATE), which was introduced in v5.4.0.

The NGATE module, which produces near LIDAR quality terrain models from optical imagery, has been shown to reduce editing time by more than 30%. Numerous new tools for the Interactive Terrain Editor (ITE) module increase productivity for creating bare-earth terrain models from NGATE and LIDAR data. In addition, enhancements to SOCET SET’s Feature Extraction (FE) and SOCET for ArcGIS® (SFA) modules have been implemented based on customer requests. For example, automatic height attribution includes an option to measure a point to set height attributes when terrain data is not available. Moreover, SOCET SET’s unrivaled provision of advanced sensor models continues with new models for the NextView satellites, WorldView-1 and GeoEye-1, as well as ALOS, EROS B, and FORMOSAT-2. Further productivity improvements have been made throughout the SOCET SET workflow. All of these new capabilities reduce labor hours through optimized, end-to-end workflows, resulting in significant cost savings throughout the mapping process.

December 2007 | Tips and tricks

Zooming screenshot tips

Interpolation Methods

Pixel interpolation is the process of determining a value for an unknown or new pixel, based on known, and typically nearby, pixel values. SOCET GXP supports four pixel interpolation methods for zooming and for generation of reduced resolution datasets (RSets). When selecting an algorithm for use, consider the time constraints and image quality requirements of the products to be created.

The following are the supported methods in highest complexity (slowest) to lowest complexity (fastest) order: lagrange, bicubic, bilinear, and nearest neighbor. The term complexity generally corresponds to image clarity; however this depends greatly on the raster dataset being interpolated, and the type of analysis to be performed.

Guidelines for use:

  • Lagrange provides the best overall and best zoom-in quality
  • Bicubic is a good choice when zooming out, but upon zooming in, the smoothing tends to blur the image
  • Bilinear is a good option to use instead of the bicubic method if smoothing becomes a problem
  • Nearest neighbor should be used for speed and performance only, as the quality results in rough, staircase edge artifacts
  • Since bilinear is hardware accelerated, it provides good performance and decent quality
  • The size and shape of a feature in the image does not matter, use any interpolation method
  • For panchromatic or multispectral images, use any interpolation method
  • For SAR data, lagrange is ideal to use with SOCET GXP
  • More artifacts are evident in images with high contrast
  • RSet generation is performed once, therefore, it is worthwhile to use a more complex interpolation method if a group of analysts will be spending time zoomed out on an image
  • Zoom interpolation takes place each time you zoom in or out on an image; a faster algorithm is generally desirable
  • Remember, what you see visually is a tradeoff between processing time and quality

The default RSet generation interpolation method is nearest neighbor. For better quality, change the interpolation method to lagrange for RSet generation. To set preference, from the Workspace Manager, select Tools > Preferences > Load/Unload > RSet Generation, or select b. You can also change the interpolation method from the Multiport main menu; select View > Zoom > Zoom Interpolation.

Interpolation method examples showing zoom
64:1 (1.562% zoom) lagrange

64:1 (1.562% zoom) lagrange

800% zoom lagrange

800% zoom lagrange

64:1 (1.562% zoom) bicubic

64:1 (1.562% zoom) bicubic

800% zoom bicubic

800% zoom bicubic

64:1 (1.562% zoom) bilinear

64:1 (1.562% zoom) bilinear

800% zoom bilinear

800% zoom bilinear

64:1 (1.562% zoom) nearest neighbor

64:1 (1.562% zoom) nearest neighbor

800% zoom nearest neighbor

800% zoom nearest neighbor

Interpolation method examples showing RSet generation
Whole scene zoom - lagrange

Whole scene zoom - lagrange

Whole scene zoom - nearest neighbor

Whole scene zoom - nearest neighbor

Customer and partner spotlight | December 2007

The U.S. Geological Survey (USGS) Astrogeology team employs photogrammetric technology to study and map geologic features of the solar system

The USGS is the primary producer of digital and hardcopy planetary maps in the U.S. Its
Astrogeology team, based in Flagstaff, Arizona, studies the properties and evolution of
planets and their satellites, asteroids, and comets. Using images and samples collected
by spacecraft expeditions, USGS researchers analyze and measure soil, water, terrain,
atmospheric conditions, and other properties that characterize solar system objects. Results
of these studies are used to construct topographic, geologic, and thematic maps, globes,
digital elevation models (DEMs), and 3D flythrough simulations to support geophysical
studies, education, and the planning and operation of subsequent missions.

Global topographic map of Mars Topographic model of Mars

The USGS produced the global topographic map of Mars, left, from nearly 1000 Viking Orbiter images by conventional analog photogrammetric methods. A team of analysts labored for most of the 1980s to complete this product, which has a post spacing of 1 km, though most of the points are actually interpolated from widely spaced contours. The topographic
model on the right, based on a single High-Resolution Imagining Science Experiment (HiRISE) camera stereo pair, contains approximately the same number of height points, spaced 1m apart. Using SOCET SET, its production required 60 hours of automated matching and 90 hours of interactive quality control and editing. The model shows debris flows associated with the central peak of the Mojave crater. Modern digital processing methods, including SOCET SET, make it practical to process gigabytes of image data for an individual research project.

USGS staff are involved in all stages of the planetary exploration and mapping process — in many cases they help to design, build, and test innovative new camera systems. Once images and supporting data are safely on Earth, they must be catalogued, processed into map products, and delivered to the customer. Ultimately, products go to NASA’s Planetary Data System, where they are archived for future users, but the immediate customer is often a team running another mission, urgently in need of maps to plan its next day, week, or year of exploration. Flexibility and efficiency of the mapping software are therefore key concerns.

The USGS has chosen an approach that makes synergistic use of both public domain
software written in-house, as well as commercial photogrammetric software. It uses
BAE Systems’ SOCET SET to accomplish its planetary mapping tasks, and its own system,
Integrated Software for Imagers and Spectrometers (ISIS), provides an end-to-end capability
for processing planetary images into orthophoto mosaics for use as base maps. By writing
the software in-house, the USGS maintains the flexibility to read, decompress, calibrate, and
model data geometrically from each new sensor as it becomes available.

Read the full case study >>