3 ms·
In case somebody is interested, this [1] is the Microsoft Research paper that introduced this schema under the name Hierarchical Triangular Mesh. Among other th
by danbruc 8y ago
In case somebody is interested, this [1] is the Microsoft Research paper that introduced this schema under the name Hierarchical Triangular Mesh. Among other things it provides an analysis of the distortion and discusses many operations one may want to perform with such a grid with a focus on GIS applications.
And in case someone is interested in the Spherical Fibonacci Mapping mentioned in the article, here [2] is a non-broken link to the paper.
[1] https://www.microsoft.com/en-us/research/wp-content/uploads/2005/09/tr-2005-123.pdf https://www.microsoft.com/en-us/research/wp-content/uploads/...
[2] https://docplayer.net/40493580-Spherical-fibonacci-mapping.html https://docplayer.net/40493580-Spherical-fibonacci-mapping.h...
- acemarke 8y agoI actually used HTM in a project a few years back. I had a 360x180 sphere of values, and I needed to rotate the sphere in some combination of translations. I did this by converting the az/el/value triplets to Cartesian XYZ and multiplying by a rotation matrix. However, when I tried to convert the translated values back to a rotated az/el/value result, the results broke due to what was effectively lat/lon narrowing (ie, 1x1-degree cells are smaller nearer the "poles", so the rotations didn't map evenly). After research, I settled on using HTM. I was able to borrow the C code from some MySQL extension that implemented HTM (looks like it's at https://github.com/smonkewitz/scisql https://github.com/smonkewitz/scisql ), then use that math to more reliably convert the rotated values back to an appropriate az/el/value result. Still one of the neatest problems I ever solved.