Composite Image Service Provides Unique Solution to GIS Developers and Map Makers
By Mike Dolbow, News Editor
When Metropolitan Council GIS developer Matt McGuire got tired of pointing his web applications to the latest aerial imagery each time a new year was added, he turned to MetroGIS for a solution. In November of 2008, MetroGIS hosted a forum where application and web service needs were defined by multiple organizations, and Matt’s “Best Image Service” was identified as one of the high-priority needs. While the name “Best Image Service” led to quite a bit of initial confusion, the end result meets the original need and more.
“The confusion all stemmed from the name,” says McGuire, conceding that the “best” imagery at any one time completely depends on the application. When he formed a workgroup that would commission a project to meet the need, much of the initial discussion revolved around this fact - and the fact that MnGeo’s Geospatial Image Server already provided dozens of choices to users. Many workgroup members had overlooked the more detailed description of the idea, “Best available image service so you don’t have to choose between layers” (emphasis ours).
What McGuire and other developers needed was a single web service that automatically switched to an appropriate set of imagery depending on scale and extent.
What Imagery is Appropriate at What Scales?
Once the overall goal was settled, the workgroup set out on the more difficult task of attempting to define what set of imagery would be “best” at certain scales and extents. The only “slam dunk” sets of imagery were true-color Landsat scenes (see Figure 1) for small scales and the most recent ortho-corrected Farm Service Agency NAIP imagery for medium scales. The NAIP imagery would also serve as the “default” large-scale background in areas that did not have higher resolution imagery available. Beyond that, the workgroup discussed a long list of imagery concerns and agreed that decisions about large-scale, high-resolution imagery choices would primarily come down to balancing three: currentness, resolution, and accuracy. Given that no fool-proof formula seemed obvious to balancing these concerns, the workgroup decided that a yearly meeting to consider new additions would be adequate, and formed an initial definition for the first iteration of the service.
Figure1: True-color Landsat scenes used in the Composite Image Service at small scales, such as statewide or region-wide.
The “Composite Image Service” is Born
With that definition in hand, MetroGIS set out to contract with MnGeo to develop the image service, now being called the “Composite Image Service” to minimize confusion. MnGeo already had much of the required knowledge, imagery and technology in-house, making them an ideal candidate to develop and host the service. McGuire tapped the workgroup members and MnGeo staff to acquire the data and fine-tune its display, resulting in service completion in Fall of 2010. Astute observers may have noticed the Composite Image Service in the background of MetroTransit’s interactive bus map (see Figures 2 and 3).
Figure 2: Medium-resolution photography from the Composite Image Service, as seen inside MetroTransit's interactive map.
“We’re using it in all our new web applications,” says McGuire, “it’s much more straightforward than having users pick from a long list of imagery years.” And given the statewide nature of the Landsat and NAIP imagery, the service can fill similar needs for applications outside the metro. Like all of MnGeo’s WMS services, it can be used inside ArcGIS, Google Earth, or many other GIS software packages - when the user changes the viewing area inside the application, a bounding box is sent to the service, which then returns the appropriate image, depending on the scale and extent. Even without much advertising, the service has been accessed over 20,000 times per month since February.
Figure 3: High-resolution photography centered on the area shown in Figure 2.
Sometimes, what a developer or a map maker wants is something simple to use, something that makes their life easier and “just works” without having to make a lot of decisions. By implementing a complex combination of multiple imagery sets and making some decisions up front, the Composite Image Service has made that option available.
For more information on how to use the service, see MnGeo’s Composite Image Service webpage. If you have imagery that you would like to suggest or contribute, or you would like to participate in the workgroup that will govern future additions, contact Matt McGuire at firstname.lastname@example.org or (651) 602-1964.