Custom Raised Relief Maps

Discussion in 'Feature this!' started by rosen21, Mar 20, 2017.

  1. rosen21
    rosen21 New Member
    I made a tool a while ago which lets you create custom raised relief maps. The basic idea is that you draw a box on a map around a region of interest, choose the dimensions, and then the tool creates an STL file and uploads it to Shapeways. There are lots of 3D models of famous places, but I wanted to build a tool which lets anyone choose their own favorite place and get to understand its topography. I'd love to get some more feedback on the tool, and see what uses other find for it.

    For those that are curious, this began as a project to try and intuitively visualize a particular watershed. Having an object that you could hold in your hand and look closely at from any angle seemed like a natural approach. I hacked things together in Python to build a single model, but soon realized that aside from the latitude and longitude points for the bounding box, this was actually a generic tool for creating 3D models from topographic data. So, then I tried to learn web development and scraped together a basic tool that let others create their own models.

    One of the unexpected lessons from this project is that when you zoom out far enough, the earth is actually really smooth. The first few models I made had the xy-scale equal to the z-scale, and to my puzzlement they showed almost no topographic detail. I was convinced it was a bug, until I did the math and realized that anytime you look at a large enough region, even the world's tallest mountains don't really stick out. For example, if you were to look at the entire Indian subcontinent (about 2,000 miles from North to South), even something as tall as Mt. Everest (about 5 miles in elevation) would only be a blip in any raised relief model. It's always refreshing when what you thought was a bug in your code, forces you to realize something about the nature of the real world, instead.

    Here is a little GIF which outlines the basic flow. (It even works on mobile!)


    Here is an example of the printed product:

    Last edited: Mar 20, 2017
    pendarestan likes this.
  2. stonysmith
    stonysmith Well-Known Member Moderator
    There are a few mountains where it would make sense to have the bounding box at something other than straight north/south/east/west.. some ability to turn the selection at an angle might be helpful.
  3. rosen21
    rosen21 New Member
    Definitely. I've looked into this a few times, but found that the Google Maps drawing tools don't natively support it, so the additional effort is significant. I'll probably add support for printing arbitrary polygons which should resolve these types of issues. Thanks for the suggestions.
  4. MrNibbles
    MrNibbles Well-Known Member
    If my math is correct even the Mariana Trench at ~10,910 meters deep is only about 1.7 percent of the radius of the Earth. The planet is rather smooth on a global scale!

    Another fun thing for this would be a capability to add GPS waypoints to the print or manually draw a line, for example to show a hiking path that was used to transit the print of the map, but that might require a color print to show the path. For a regular plastic print maybe a shallow raised ridge or groove could be added.
  5. stonysmith
    stonysmith Well-Known Member Moderator
    A common billiard ball, as smooth it is, would have deeper/higher terrain than the earth does, if they were the same size.
  6. MrNibbles
    MrNibbles Well-Known Member
    That depends a bit on whether you consider the production specs for billiard balls to be more related to smoothness or roundness. The diameter of a ball is specified as 2.25 inches plus or minus 0.005 inches, or basically +/- 2.2 %. But in real life a ball is polished smooth to a much greater degree than the diameter allowance, and measurements of diameter excursions generally don't measure surface roughness unless they are gouges or ridges wider than the caliper jaws or other tools that might be used. Earth is also an oblate spheroid because it's spinning and that complicates things even more, as compared to a spherical billiard ball. A better review of billiard ball production standards might reveal whether smoothness and "sphericity" have individual specifications.

    So really there are two measures of concern here. One is a spherical measure and the other is roughness or smoothness measure. To compare "smoothness" we would need to dig up RMS measurements of the surface deviations of both a polished billiard ball and satellite measurements of the planet, probably supplemented by underwater sonar measurements. Or we could just consider RMS measurements of land masses above sea level.

    There's probably an analog here to how Shapeways specs their material tolerance. You could print a large thing that ends up being terribly warped and it would be way out of whack with their tolerance specs. But at a smaller localized scale tolerances are determined by stepped print ridges, mechanical polishing, etc. an that's the one that Shapeways is talking about.
  7. stannum
    stannum Well-Known Member
    11 / 6371 = 0.001726...
    ~0.17 %
  8. MrNibbles
    MrNibbles Well-Known Member
    Eh, that damn metric system always gets the better of me.

    An interesting effort is the Avogadro project seeking to make mass standards out of silicon spheres, and apparently they are polished to a degree that if the spheres were scaled up to the size of the Earth they would have a smoothness with deviations of under a few meters. And of course they are much more spherical. A similar thing was done for a gravity relativity research project in a satellite but with smaller polished spheres.