By
Paul G. Garland, vice president,
Z/I Imaging Corp., Huntsville, Ala. (www.ziimaging.com).
Many mapping
professionals are familiar with a variety of Earth imaging software
tools that offer advanced capabilities for image processing, feature
classification, and feature recognition and extraction. Some are
enhanced to handle multispectral and hyperspectral imaging, and more
information is being derived from data-fusion techniques.
Many of these functions have boundless benefits, but are time consuming
and costly. They become even more complex when used in a production
workflow environment, requiring easy-to-use cataloging and image
management software to achieve optimal productivity. Such products
include EMC’s Image Storage Architecture, Leica Geosystems’ GeoVault
Data Manager, Red Hen Systems Inc.’s PixPoint and Z/I Imaging Corp.’s
TerraShare.
Enterprise Data Management
Effective data-management products query and search massive amounts of
data, taking the guesswork out of finding and managing information.
Production shops need help managing data. But the data are linked
directly to production efficiency, and tracking data isn’t enough.
Process-management tools must be tightly integrated with data management
products so managers can effectively track product generation, employee
performance, product quality and, most importantly, project cost.
Effective data management software manages and tracks data throughout
the enterprise. Some product offerings have evolved into geodata
management systems that manage images, elevation data, vector
data–basically any data related to a physical location. Such
capabilities extend beyond traditional cataloging systems; they include
a programmable ability to understand geospatial data content and extract
metadata (number of rows and columns, compression techniques and
parameters, number of pyramid levels, etc.). In addition, some products
store user-defined metadata such as acquisition date, operator
identifiers, miscellaneous data, process logging and other
workflow-related information.
With enterprise data management software, users aren’t concerned about
data location; they access data through logical files and folders or
geospatial location, not through arbitrary or confusing file names. Data
management systems offer different user interfaces. For example, Red Hen
Systems’ PixPoint software integrates directly into ESRI Inc.’s ArcGIS
package (see Figure 1). Users access image data through the geospatial
view in the ArcView client, or through a thumbnail window.
Z/I Imaging’s
TerraShare product extends the Windows Explorer user interface. Through
motions as natural as file operations in Explorer, users manage data,
create catalog entries, enter metadata and user metadata, view data and
manipulate data within the TerraShare environment. Through natural
Windows methods, users interact with TerraShare through drag and drop,
cut, paste, copy, move, search and delete operations. Advanced functions
(like SQL Query capability) make searches—based on user-defined
metadata—as easy as finding a file in Explorer. For example, a Microsoft
Word document or scanned legal paperwork can be entered into
TerraShare’s storehouse and retrieved by queries on metadata, geospatial
location, data type or file name. A geographic view of data can show the
document tied to a single location
and area, or several locations. By double-clicking or right-clicking on
the document, users can launch Word or any associated application on the
georeferenced data (see Figure 2).
In the more
advanced data management systems, a “logical” file is really one or more
physical Windows files. Using this capability, users can refer to
multiple files with one name, moniker or footprint, and applications can
operate on complex file clusters that users conceptualize as a single
entity. For example, a hyperspectral image may be composed of 256
different files, each one a single-band JPEG2000 image. The system would
treat them as a single entity—to be processed and managed as a single
image—with common metadata, but possibly dissimilar extended metadata.
Geodata management software products are relatively new, and there’s
plenty of room for improvement. For example, certain common extended
metadata (workflow statistics) are used by most production shops. Data
that relate to the current production status of products being developed
are typically carried along with the product itself. Information such as
percentage complete, operator name, operator cost, estimated time to
complete, estimated cost, actual cost, etc., helps control production
costs and validate customer expectations.
Workflow Management
The next great productivity leaps will come from a new direction in
workflow enhancement: workflow management. Workflow management
encompasses several concepts: process management and monitoring,
user-defined data visualization, distributed and queued processing, and
programmable services. Such functions will allow users to manage
workflows and geodata. Together, they will improve users’ workflow and
production management capabilities by automating interprocess functions,
improving process flow monitoring capabilities, increasing CPU use, and
improving and standardizing graphical data and process status views.
Combining geodata management and workflow management into a
user-friendly, efficient system will increase production throughput and
lower product costs.
Extending
workflow management principles to the enterprise means taking advantage
of enterprise data management, process management and process
monitoring, data visualization, distributed processing and enterprise
programmability. Clearly the largest gains from these attributes of
enterprise workflow management can be seen in distributed processing and
enterprisewide access to data through the next generation of geodata
management software.