TNRIS: Change Detected

At the TNRIS GIS Forum, Richard Wade gave a sneak preview of TNRIS’s new web site, scheduled to go live in December. I was impressed. It showed an easy to use drag and drop interface allowing easy mashups of SqlServer08 spatial data with Virtual Earth.

New TNRIS Website – Designed by Neogeographers
Richard delicately pointed out that the new web site was designed by web developers – not GIS Professionals. Reading between the lines, I sense that TNRIS wanted to see if spatial really is that special. From what I can tell, the moral to the story is that it’s easier to train new web developers spatial skills than to teach GIS developers new web skills.

Still, I think there are places where spatial is special. Let’s take a look at the backend. All the work that goes into preparing models to produce timely results that can be mashed up for the world to see requires special skills. Look at the image currently on the front of TNRIS site …


It appears to be pre-Ike. As post-Ike imagery becomes available, many people will want to see change detection analysis results. It would make life easier if there were a site where I could upload change detection programs and run them, without having to download the imagery locally, run my program, and then upload my results.

Unsure about Azure
Until Azure is available though, the only option for scaling out a .NET program is by writing a lot of complicated code to scale out across multiple processors. Scaling across multiple processors on the same machine is quite a different challenge than scaling across multiple machines in the cloud. Azure is limited to one processor per machine, so there’s no point in writing multiprocessor code for it. Unlike EC2, VMs will not share machines. That means code written to scale across multiple processors is useless if I end up going to the cloud. In Azure, scaling out involves use of the Fabric Controller. I don’t see much info yet on the Fabric Controller, but I suspect it will be a lot easier interacting with it than with multiple processors on the same machine.

2 comments so far

  1. Jarlath on

    Change detection of high resolution imagery could really benefit from being moved to the cloud. Pixel-based methods of change detection are not very effective when the resolution of imagery is better than 5m, requiring the use of object-based feature extraction. Object-based feature extraction is wonderful technology, but it is a resource hog, required data to be tiled, processed, then mosaicked.

  2. Kirk Kuykendall on

    The NAIP imagery is 0.5m. I could see how generalization will require a lot of processing. Seems like a WeoGeo approach whereby producers of “derived products” are incentivized to make their results available to others would be a good thing.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

%d bloggers like this: