Wind Power Classification: Relief Map in Augmented Reality

After working on my first experiment with Augmented Reality I decided to try and apply the technique to help re-visualize an existing data set.  I’ve always been a big geek for renewable energy, especially solar and wind power, so I selected the NREL’s Wind Power Classification map as my test subject.  The result can be seen below:

AR relief map of projected United States wind resources

AR relief map of projected United States wind resources

Creating this AR visualization was a bit of a challenge, as I was unable to rely solely on simple geometric primiatives as I did in my Earth-Moon system demonstration.  Here’s a rundown of my process:

  1. I took the NREL Wind Power Classification map and extracted the map portion in Photoshop.  I then took this reduced version of the map and brought it into Flash, where I transformed it into a vector object.
  2. Next I re-keyed the colors on the map to utilize a more distinct palette.  Colors used in the visualization are constructed using combination of red, green and blue as follows:
    • red:  100% R, 0% G, 0% B
    • blue:  0% R, 0% G 100% B
    • purple:  100%, 0% G, 100% B
    • green:  0% R, 100% G, 0 % B
    • yellow:  100% R, 100% G, 0% B
  3. I then wrote a quick program in Processing that would open up this re-keyed image and perform color analysis on each pixel.  The program assigns values to each color (blue = 70, red = 60, purple = 50, green = 40 and yellow = 30) – once the color values have been established the program takes a second pass and averages the colors within 10×10 regions and outputs these values to a text file.
  4. Next I created a basic AR movie in Flash.  I imported the re-keyed map and applied it to a Papervision 3D plane object.  Plane objects in PV3D can be constructed to contain verticies, so I set up my plane to have 50 x 31 individual verticies to match the output of my Processing sketch.
    Plane object without applied texture - note the individual verticies!

    Plane object without applied texture – note the individual verticies!

  5. Finally I wrote a routine to bring in the text file and adjust the height of each of these verticies based upon the average wind power data as determined by the algorithm in the Processing sketch.

If you’d like to try the AR yourself feel free to do the following:

  1. Click on this symbol and print out a copy
  2. Click here to access the Flash video
  3. Click the ‘Start’ button
  4. When prompted, let Flash access your webcam
  5. Hold up the symbol to your webcam to view the simulation
  1. Pingback: EDAD660 » Augmented Reality

  2. Tommy
    Hello We are a group of students from Denmark who are doing AR as a student project... we have got the AR basics working using FLARToolkit and are now looking to get some motion and animation into our application... We were wondering if You would care to share your source code for the Solar system application you made, as wecas see it contains some movement (rotation of the planets etc.)... the source code, if you are willing to share, can be mailed to: meab347@gmail.com - Thanks
  3. Paulina
    This really is incredible, but I have no idea how that flar-logo marker translates into an image. I have some learning to do. Fascinated.
  4. Wind Power Australia
    Nice post. Looks like wind power is really starting to get some serious consideration in Australia now.

Leave a Reply


captcha *