For each brick that has dataset points assigned to it, determine if that brick should be land or water. There are many ways to do this -- I used simple voting of whether each point was land or water. I also tried different thresholds and weighting each "vote" by the corresponding altitude or sea depth. Write this dataset of brick coordinates to a file.
The program READGIS was written to perform these operations. It reads in a geophysical datafile and, for a globe of a given size, calculates the color and 3D coordinates of the bricks which will make up the surface of the globe.
Using
the dataset of brick coordinates, 3 programs were written to generate and
display the various 2D and 3D images that were created.
These programs are listed below. Click on the image to be taken to
the appropriate image gallery for each program.
|
DRAWGLOBE this program draws all of the bricks in the dataset in 3D space. creating the various 3D images shown on these pages. Keyboard commands allow zooming in and out, rotating in both the phi and theta directions, and writing the currently viewed image out to a PNM file, a standard bitmap format which can be converted into many other formats by various graphical software packages. I tried to integrate some rudimentary ability to edit brick colors with the mouse, but it didn't work too well .... converting an X,Y mouse coordinate to a projected x,y,z point on the surface of the globe and also considering which type of 3D projection OpenGL was using was rather nontrivial .... |
|
FLATGLOBE this program takes all of the bricks in the brick dataset and projects them onto a 2D latitude/longitude image (like the original data set). This image provides an idea of how well "sampled" the original data is -- the latitudes towards the equator are represented with better resolution than the latitudes towards the poles. Obviously, the larger the radius of the globe, the better the sampling. This program also allowed panning and zooming, as well as toggling the color of individual bricks, writing the modified dataset, and writing the displayed image to a PNM file. |
|
|
CROSSSECTION You guessed it. This program views horizontal (z plane) crossections through the brick dataset and can write them to individual files. These are used for the actual construction of the globe. |
The biggest challenge to building the globe was structural stability - it should not implode or fall apart when handled! The Master Builders at Lego® use glue for their creations, but that takes half the challenge out of it! Another challenge is getting started with the initial construction, since the base has a rather small diameter at the bottom and then rapidly expands outward.
Before I had finished writing
the software (described above) to determine the colors of the surface elements,
I started building prototype spheres of smaller sizes. I started
at the bottom (south pole) and built upwards. In general, I found
that for sufficient structural stability, you need the internal structure
of the first 3-4 levels of bricks to be relatively solid. After that,
a cross-like structure which rises vertically through the center of the
globe is rather rugged, and provides a decent base upon which to expland
outward for support when completing the top 4-5 levels of bricks. Here
are two pictures of the half-completed globe:
The final creation is
28.8 cm in diameter (36 brick widths, 30 brick heights, 11.3 inches).
I estimate that it utilized
the eqivalent of about 4 of the 1000-or-so brick "brick-buckets", however
a lot of the smaller (1x2,1x1) bricks went unused. Here is a picture
of the final creation. Click on the photo to be taken to the gallery
with more pictures.
NEXT
Resources (where I got the data set, software used, etc)
BACK
to Why I did this
BACK
to Planet Brick
BACK
to Rob Butera's Home Page