Feature: Workflow Integration at the National Geodetic Survey
Implementation of a new workflow solution allows the NGS to easily manage multiple, large projects with fewer resources.
By Stephen White and Steve Riddell
The National Geodetic Survey (
NGS), a program office of the National Oceanic and Atmospheric Administration’s (
NOAA) National Ocean Service (
NOS), maintains a coastal mapping program with the goal of providing accurate, consistent, and up-to-date national shoreline data for the United States and its territories. The NGS is responsible for conducting remote sensing surveys of coastal margins of the United States and its territories for demarcating the nation’s legal coastline. NGS also creates the baseline for establishing the nation’s territorial boundaries and Exclusive Economic Zone.
The Integrated Ocean and Coastal Mapping Mission
NGS and its predecessors have been conducting shoreline mapping activities since the original ‘‘Survey of the Coast’’ in 1807, and the shoreline depicted on NOS nautical charts is treated as the legal shoreline by many U.S. agencies. The shoreline data generated through these surveys are incorporated into nautical charts to facilitate safe marine transportation and navigation.
Currently, NGS extracts and attributes shoreline data from tide-coordinated stereo photography in a softcopy environment. Tide/time windows are established for tide-coordinated photography based on tolerances recognized in NOS guidelines for mean high water (MHW) and mean lower low water (MLLW). In addition to its primary use on nautical charts to assist in safe navigation, this national shoreline serves numerous other purposes, ranging from determination of legal boundaries to coastal management and environmental applications such as climate change studies.
In another effort to support NOAA’s homeland security and emergency response requirements, the NGS can acquire and rapidly disseminate a variety of spatially referenced, remotely sensed datasets to federal, state, and local government agencies, as well as the general public. Over the last several decades, NGS has assisted with recovery efforts following a variety of natural and human-induced disasters, including earthquakes, tsunamis, hurricanes, nor’easters, and oil spills. NGS plans to acquire remotely sensed data in the future to support the agency’s homeland security and emergency response requirements. This includes providing tools, technology, and expertise in a timely and efficient manner during an emergency response effort.
The data will be disseminated to facilitate support efforts such as, but not limited to:
- aiding emergency managers to develop recovery strategies,
- assessing damage through comparison of before and after imagery,
- rebuilding damaged properties, and
- allowing those displaced to see images of their homes and neighborhoods.
Remote sensors (e.g. digital cameras) can provide data in a quick and efficient manner for use in emergency response efforts. Digital cameras with direct georeferencing capabilities have allowed NOAA to greatly facilitate its emergency response efforts. This technology allows high-resolution digital imagery to be made available for emergency personnel and the general public in a georeferenced GIS format, within 24 hours of acquisition, through the Web. Perhaps most importantly, the data for both mapping programs can be collected in a manner to support a variety of Integrated Ocean and Coastal Mapping (IOCM) applications, such as nautical charting, storm surge/tsunami modeling, coral reef mapping, ecosystem monitoring, and coastal mapping.
A primary objective in NOAA is to develop operational procedures to support IOCM. This effort is to efficiently and cost-effectively collect and distribute data meeting a variety of coastal mapping needs across NOAA and the rest of the federal government. The purpose is to decrease the duplication of mapping efforts along this margin through acquiring data once and having several groups use that data for their needs. Therefore, the digital imagery used in this program can be used through an IOCM effort to increase both the quantity and quality of information about the coastal environment.
Goals of the Integrated System
The digital camera imagery collected for the NGS mapping programs is acquired in a manner that allows for the creation of georeferenced, mosaicked orthophotos. Although in the past this has not been a standard deliverable, the derivative product can be used by others for numerous geospatial applications that NOAA has possibly not envisioned.
The workflow to produce this product uses a blend of in-house-developed software and commercial off-the-shelf applications. A geospatial process management system was identified as an integral component needed to support, organize, and manage the creation of this NGS deliverable. This process management system will allow for productivity gains and improvements, through wrapping all of the diverse software packages currently used into streamlined graphical workflows, organized procedural processing steps, and the capability to distribute processing to several workstations. The mosaicked orthophotos derived from this system are made available through NOAA’s Digital Coast (
www.csc.noaa.gov/digital
coast/) to other federal, state, and local agencies as well as the general public.
Workflow Management Solution
The workflow solution the NGS selected for the IOCM imagery system is by
GeoCue Corporation. GeoCue had previously deployed a solution for the NGS lidar production system, and, thus, the imagery solution forms a natural extension to NGS’ evolving vision of a comprehensively integrated workflow.
The IOCM data-processing workflow requires a sequence of processing through disparate software applications. The data sets include raw images (a few thousand to tens of thousands), elevation surface models, camera exterior orientation for each image, and camera calibration data. The processing applications include
Trimble’s DSS ImageView for raw image development (that incorporates an Inpho orthorectification module),
Inpho’s OrthoVista application for radiometric correction, tone-balancing and mosaicking, and other custom utilities for generating metadata for the final product delivery.
A workflow management solution was developed to integrate the individual applications and to efficiently manage the data and processing. GeoCue’s commercial, off-the-shelf geospatial process management system was selected as the basis for the system because it had been successfully deployed in the lidar production system and would allow integration across these two workflows.
Additional customization was included to integrate the particular applications into a GeoCue CuePac, which is a customized environment in GeoCue for a specific type of workflow. This customization includes creation of checklists that specify the sequence of processing steps as well as provide user-interface elements for invoking the applications for performing the particular steps. Command line interfaces to the applications allowed simple integration into the GeoCue process management system.
Processing
The processing starts with an exterior orientation file (provided from post-processing
Applanix Position and Orientation System data), which includes individual exposure locations. In addition to specifying this file, a named calibration set is selected from a list provided by the GeoCue Camera Manager. Because NGS owns multiple DSS cameras and each camera has current and previous versions of calibration data, it is important to have a system for ensuring that correct calibration data is selected for a given camera and collection date.
Once the user specifies the input image files, elevation data, and calibration information, a graphical representation of the project is created. Footprints indicating the ground coverage of the exposures are then viewed in a map window, which can include other vector or raster reference data and can be used to quickly confirm the coverage of the flights.
Data files are automatically managed, and you can select and process them by graphically selecting elements from the map window and activating the associated checklist steps. This greatly enhances the robustness and efficiency of the processing flow.
After project setup, exposures are selected and a Set Parameters step is run to set the desired image “development” and rectification settings. These settings are stored as properties on the footprint entities. Simply selecting a footprint again, at any time, allows you to review the settings in the standard GeoCue Properties pane. This is the standard form used throughout GeoCue for managing spatial objects such as images, elevation files, and other geospatial data.
Once the user enters the settings, he or she starts a process step. As mentioned earlier, constructing checklists is part of the customization of the workflow. Some steps require user interaction, while others can be run in a batch or distributed in processing mode. The processing mode is part of the logic built into the checklists of processing elements. For image development and rectification, the Process step sends the images to as many machines as are available and configured for this type of processing.
The management of processing across local processing “clouds” is orchestrated by the GeoCue Command Dispatch system. It includes the Dispatch Manager, a set of tools for monitoring and managing all the computers in the system as well as showing all the tasks scheduled, running, or completed. Computers can be taken on- or offline and even scheduled to be online during certain hours and unavailable at others. An important capability is a priority preemption system that allows a high-priority task to delay a lower priority task, even if that task is already in progress.
As the image development and orthorectification completes, the footprints are filled in with the image data in the GeoCue Map View. The following step is to mosaic the orthorectified images into a regular grid of products, uniform in size and orientation. Built-in gridding tools make this easy, including specialized options for naming the products and setting optional border distances (allowing the creation of overlapping mosaics).
As with the orthophotos, as the mosaic products are completed, their footprints are filled with the resulting image data. A final step generates project metadata, taking its inputs automatically from properties on the GeoCue project and images.
Since the installation of the GeoCue-driven IOCM system in the first quarter of 2011, more than 50 projects have been processed through the workflow, with more than 12,000 ortho mosaics having been produced and integrated into the NOAA Digital Coast. These data represent more than 17% of the optical imagery data downloads from Digital Coast. The encapsulation of the ortho production process into a workflow environment has allowed the NGS to easily manage multiple, large projects with limited staff.
Stephen White is a remote sensing specialist within NOAA’s National Geodetic Survey’s Remote Sensing Division. As a member of the research and development team, he works on projects that involve evaluating new remote sensing technologies/systems for integration into NOAA programs, such as the Coastal Mapping Program and Airport Survey Program.
Steve Riddell is an engineer and software developer for GeoCue Corporation in Madison, Alabama. He has more than 16 years of experience creating software for photographic and lidar imaging systems for the surveying and mapping industry.
» Back to our Aerial Mapping Spring '12 Issue