Tuesday, January 09, 2007

Google Earth as a Geographic Information System

The first work that I did with Google Earth and KML was to build a converter in ArcMap that would export all of the selected layers out to KML. We used this for an emergency response tool set that worked in ArcMap. The idea was that the mapping would be managed by someone trained in using GIS who would them automatically kick-out KML updates to a server to be viewed using GE networklinks by commanders.

This worked really well, but at the time, I wondered if it would be possible for Google Earth to act more as an interactive partner rather than a static viewer. When it came to having GIS capabilities, Google Earth has never really been considered (for good reason).

However, the more I thought about it, the more I wondered if GIS type functions used for command and control applications could not be accessed and used by Google Earth. What I really wanted to do was to used open-source functions with not license restrictions to create a GIS back end, and then have Google Earth act as the interface.

When Google Earth officially announced that they had a COM API, I started looking a bit more at the possibility of being able to interact more effectively with Google Earth. I noticed some update functions that I thought might help . I first mentioned this in the beginning of December when I wrote New Google Earth API Allows for GIS Type Interaction.

I was able to use this new API function to select coordinates directly from Google Earth. This has huge implications, as I can use these coordinates to Select and Identify GIS data. I can now marry this capability with Shape2Earth, which allows me to Import GIS directly into Google Earth.

So, now I can load GIS data directly into Google Earth, and I can query and select this data using Google Earth. The next thing I want to do is actually create data and launch geoprocessing functions from Google Earth. Here is how I do that.

To start with, I have the Google Earth Render window embedded in my application (I'm using GoogleEarthAirlines to do this), and I have loaded GIS data using a modified form of Shape2Earth.

To show how I can interact with Google Earth similar to a GIS, I'll walk through an emergency response scenario. I am given a building at which there is an incident, I can do a quick search using either the building name, ID, or a coordinate. I use MapWindow GIS components to search for the building information in the shapefile, and then zoom to the location.
When I get to the area of interest, I want to Cordon (buffer) off the affected area. I first select my Cordon tool and click on Google Earth to set the center point (I could also select the building and create a cordon around that). The point (or selected feature) is sent to my cordon tool. I then select the size of the cordon, the measurement units (feet, meters, etc.), and the number of rings I would like. In this case, 2000 feet with 1 ring.
MapWindow GIS components are used to create the buffer (shown below). Shape2Earth creates a kml file and a shapefile to represent the new geospatial data set.

The next thing I want to do is determine what buildings are impacted by this event. MapWindow GIS components select the buildings using the cordon polygon. Shape2Earth modifies the color of the kml and reloads is quickly into Google Earth (see below). The same process is used to select affected roads, and to set up traffic control points to isolate the area.
All of these new data are saved as kml with associated shapefiles as described in Importing GIS Data into Google Earth.

Since my GIS data and KML are linked, and I can grab coordinates off of the map, I can change the status of buildings simply by clicking on them in Google Earth (note that the orientation of Google Earth does not matter, not does the location that you click). This brings up a dialog box that lets me edit the GIS data which also changes the color of the KML.

Above, GIS data is selected and edited directly in Google Earth.
Below, the KML file is update and reloaded.

Combining the visualization of Google Earth with the power of GIS leads to some very effective capabilities. The data can be immediately served out to other clients, such as another Google Earth as seen in the graphics below.

GIS Data viewed and edited in Google Earth above

Created GIS data viewed in another Google Earth Client in near real-time

Besides being able to edit GIS attribute data, we can create GIS data by digitizing directly on Google Earth. The graphic above shows a sketched route that is converted to a shapefile.

Importing GIS Data into Google Earth

Google Earth has often been described as a geospatial data browser. The reason for this is simple ... while Google Earth is fantastic for visualization, it does not do (nor has it been intended to do) many traditional GIS tasks. As an example, you cannot really query GIS data or directly access attribute data. You also cannot edit or manage geospatial data. Google Earth is really intended to do one thing very well, and that is to display geospatial data on a visually appealing framework. There have been many efforts by a lot of people to try and wring GIS type capabilities out of Google Earth. Many of the more successful ones used a combination of network links and data embedded in the description tag (and visible in the callout balloon) to link points to other data available on servers. What would really make things easier would be if users could click on Google Earth and grab coordinates. If we could do this, then we could attach geospatial capabilities behind Google Earth to make it act as a GIS interface. The new Google Earth API (and a lot of trickery) have now made that possible. The first thing I wanted to do was to be able to create toolbars that could be accessible in Google Earth. Since I could not do that, I started looking into grabbing the Google Earth Windows handle so that I could embed the viewer in my application. I had a 75% solution in the bag when I stumbled on GoogleEarthAirlines. GoogleEarthAirlines is an ActiveX control written in Delphi that does just what I was planning to do. When I found this, I put my control on the backburner and started using this. The graphic below shows the GoogleEarthAirlines control in my application.

The plan now is to link GIS data with KML. KML will act as the layer file that does nothing more than render my GIS data. to get the GIS data into Google Earth, I used a modified form of Shape2Earth. Shape2Earth converts the shapefiles to KML, and then writes an XML file that links the two files together.

Shape2Earth is used to create height values of GIS Data

Basic colors for the output KML are selected

A Unique ID is selected from the Attribute field. This will be used to attach to databases through Google Earth

The steps above walk through loading a shapefile into the Google Earth GIS application. Now, we use our trickery to click on Google Earth and grab a coordinate. This is done using the method described here.

The great thing about this method is that it does not matter what the orientation of Google Earth is. It will always return the coordinate that you clicked.
In the graphic below, I have selected the ID button, and then clicked on a building. The application uses geoprocessing capabilities provided by the open source GIS application MapWindow. It loads the selected shapefile, and uses the coordinate to determine what has been selected. The result is the ability to directly access geospatial attribute data by clicking directly in Google Earth.

The same method can be used to select multiple features, as is shown below.

Select Tool lets the user drag a selection box across Google Earth

MapWindow components are used to select the geospatial features and display their attributes
Now that we can do some basic GIS type functions directly in Google Earth, we can start doing more sophisticated workflows. The next step is to actually create GIS data directly in Google Earth. That has already been done, and will be posted here shortly.

Thursday, January 04, 2007

Shape2Earth Version 1.0

After three beta testing periods, and a lot of new enhancements and optimizations, Shape2Earth Version 1.0 has been released for download at http://shape2earth.com/.

Shape2Earth uses freely available Geographic Information System (GIS) software to convert shapefiles into color coded Keyhole Markup Language (KML) files for visualization in Google Earth.

Shape2Earth provides a powerful tool for people and organizations who do not wish to invest thousands of dollars in professional GIS software in order to create powerful and compelling visual information. Shape2Earth can also quickly expand the capability of those who already have a significant investment in GIS for the lowest price possible.

Download Shape2Earth, and then look at the online blog to see what shape2earth can do for you.

Physiographic Regions of the U.S.