Imagery is an important component of ArcGIS 10. With the ArcGIS for Server Image extension, massive volumes of imagery files, such as satellite and aerial data, can now be accessed very fast. The demo shows how a seamless mosaic is rendered dynamically on the fly. Multiple sources a brought together and interpreted at the same time. Users can also now see changes over time by merging information from different timelines.
Jul 12th, 2010
00:01Imagery is a core part of GIS.
00:03At ArcGIS 10, we've done a number of things to improve how we interact and work with imagery...
00:08...how we can visualize it, how we can make it look beautiful...
00:12...how we manage large collections of imagery, and how we can extract value from those pixels...
00:17...more than just backdrop maps, but use that imagery to help us create features and understanding.
00:24I want you to remember three words...massive, fast, and accessible. So here we go. Gerry Kinn and Lindsay McGreevy.
00:33Thank you, John. I know many of you manage large imagery resources...
00:37...and I know from my experience that whenever you do that you're faced with three challenges...
00:43...the data's always massive in volume; there's always a diverse set of sources as well as user needs...
00:49...and, of course, time's always the factor you have to deal with.
00:53What I want to do is lead you through what Server provides to solve some of these challenges.
00:59Lindsay and I started this project by going to my home state of New York and visiting with Department of Transportation...
01:06...with critical infrastructure, and with universities to name a few.
01:09We gathered up several hundred thousand images, and we brought them back and stood up a server with several terabytes’ worth of data.
01:18Let me show you what we've got. Of course we have things like the new NaturalVue, which will be debuting soon.
01:26We have probably a coverage you're familiar with...with ArcGIS Online.
01:31But let me step in and show you the footprints for the collection areas that we have, associated with New York State.
01:40In particular, let me turn on...
01:43...not just these overviews of satellite data, of imagery from airborne, different formats, some compressed, some uncompressed.
01:54But let me turn on the footprints for one single layer. This is 50,000 images, and let's dive in and see what this looks like.
02:02What Server is doing as I turn on this layer is it's dynamically going to disk and rendering this mosaic that you see.
02:12I'll zoom in again and after I do that; let me turn off the footprints. What you see is a seamless mosaic.
02:21So it's collected all these images, dynamically rendered this and mosaicked it on the screen.
02:25What you...the result is a very intuitive natural image that you can use.
02:31Let's go to Rome, New York. I want to show you an example there.
02:36Since this is multiband data, when we zoom in to Rome Free Academy, we see two ball fields...
02:42...but as I go into the image and I turn on a color infrared layer, this is simply another view of the same image.
02:51We see that this side is natural grass and, of course, this is artificial turf...classic textbook example. Okay.
02:59Not all imagery, if you will, is visible imagery. This here is an elevation dataset dynamically rendered as a shaded relief.
03:13Now, here is standard elevation data for the state which is rather low resolution, but let me take you to another part of the state.
03:21You can blend these resolutions. This is very high resolution from processed lidar.
03:27In fact, I can even see the roadbeds here very clearly.
03:31Let me pull back this. You can see that road. But if you look here, you can see what looks like another roadbed.
03:39Yet, what we have here is a roadbed that's been overgrown by a forest. It's abandoned for some period of time.
03:46The point I want to make is these diverse datasets, when they come together in the geography, we're able to make better interpretations.
03:54We're able to step beyond what we could do easily with a single source.
03:58Okay. I want to take you to western New York now.
04:02And this is exciting because we've had several decades of Landsat continuously collecting data looking down at the earth.
04:11The layer I'm turning on is an analysis that's been done that looks at the significant human change to the terrain...
04:22...not the seasonal change, not these other things, but this is the human change to the terrain...
04:27...and it's color coded according to the year that it occurred.
04:30You can see near the Buffalo Airport, there's quite a bit of change. Let's go in and look a little closer.
04:37What I'll do is I'll turn on several years of data, first beginning in the 1990s.
04:44I want you to look at this industrial building here; the runway; and, of course, the terminal areas.
04:50As I move to 2002, you can see that this building is gone.
04:55The terminal area has been worked on, and there's some more work going on here. And by 2008, this is all complete.
05:02The point I want to make with this is you can go further with imagery. You can analyze it.
05:08You can use it to keep your GIS current...
05:11...and if you know where the change is occurring, you can concentrate and focus on those areas...
05:16...and actually become efficient enough so that you can maintain your GIS.
05:21Okay. This is a very, very quick glimpse in a Flex viewer of what Server can do in order to help you share your information.
05:30But we really expect in GIS 10 that you're going to take the next step and begin to collaborate.
05:39This dataset that we have not only has this complete sort of blanket coverage of the state, but it's also got a lot of project data in it as well.
05:51Let me turn on one specific example, and this is a DOT project in the eastern part of the state...
06:00...and we're going to have Lindsay show you that next step...
06:03...of how you can begin to work with this in a more detailed fashion at desktop. Lindsay.
06:11I've been an analyst for 17 years now and traditionally a platform with both GIS and imagery has been very challenging...
06:19...until now. With ArcGIS 10, GIS and imagery come to life.
06:25Let's focus on the Lake Champlain Bridge, which connects New York and Vermont states.
06:31Notice we have three different types of imagery, all of which come from the same seamless, statewide dynamic mosaic that Gerry was using.
06:41On top, we have the Department of Transportation collection, black and white pan; they like to clip to their project area.
06:49Underneath, we have color infrared that goes back to the 1990s.
06:54Underneath that, we have four-band natural color statewide for 2005 to 2009, all of which I can access in my desktop.
07:05So let's take a closer look at the imagery.
07:11Our image looks okay, but it could definitely look a lot better. Now how would I go about improving this appearance?
07:17Well, I'd use the new Image Analysis window built into ArcGIS in all of your desktops. The Image Analysis window is made up of two parts.
07:27The top part is a display enhancement and the bottom is on-the-fly processing...
07:32...all of which are available to you very easily and accessible with this new window.
07:38So let's take our natural color imagery and tweak it a little bit, changing the brightness, the contrast...
07:49...and actually let's go ahead and do a stretch.
07:52Now pay close attention to the image. You're going to see it improve immensely.
07:57Doesn't that look a whole lot better? All of this is done on the fly.
08:09Let's take a look at this interesting feature in the landscape.
08:13This is actually a British fort, the oldest British fort in the United States. It goes back to 1750s.
08:19You can see...I could improve it a little bit more by sharpening this image...again, all on the fly.
08:27So let's take a look...
08:31...and actually Flickr between the original and the new, sharpened image. So here's the sharpened image.
08:40There's the original.
08:45Notice you can see the detail on the parapet walls. You can see the barracks. You can see the details of the trees a lot better.
08:52And, in fact, you can actually see the cracks along the road.
08:59So now let's take a closer look at the pixels at a really good resolution.
09:04Because we have this high-res. Department of Transportation data for the project, we want to literally cruise our project area...
09:13...and I mean cruise.
09:19Isn't that great?
09:22Accelerated display is an exciting addition to ArcGIS 10. This is excellent, but it, too, could be better.
09:32How? Imagery and GIS together.
09:37I want the same performance for everything.
09:48So as you can see, in our desktop, we can make our imagery look great, fast, and sharp.