Home

Interpolations

2006, July 14 - 06:47 — Jason W Barnes

Example of 4 different interpolation styles as applied to mapping from a Cassini/VIMS Titan cube.

Comments

2006, July 14 - 09:36 — Matt Tiscareno

Re: Interpolations

Interesting. You're going from what resolution to what resolution? Shouldn't "nearest neighbor" be the raw data? Why aren't the pixels square? What do the colors represent? I've not thought much about interpolation methods, though I have a sense that my simple reprojection software is inferior because it doesn't employ the Sampling Theorem. Any thoughts on which one is best?

2006, July 14 - 10:09 — Jason W Barnes

answers

> You're going from what resolution to what resolution? Shouldn't "nearest neighbor" be the raw data? Why aren't the pixels square?

These questions go together, and I probably should have pointed this stuff out somewhere earlier. Here I went from the original, 64x64x352 VIMS cube to a cylindrical map projection at a few pixels per degree (don't remember the exact value). So, the pixels are square, as viewed from Cassini when they were taken. However, when shown in cylindrical projection, their shape is a function of that viewing geometry projected onto Titan.

> What do the colors represent?

Colors are mapped within three of Titan's infrared atmospheric spectral windows: Red is 5.0 microns, Green is 2.0 microns, and Blue is 1.26 microns.

2006, July 17 - 16:29 — Anonymous

Re: answers

Gotcha. I was going to say that, if your colors were band depths, you would want to interpolate and reproject each channel separately, and then compute your band depths. But on second thought, I think I was wrong, and you would want to compute your band depths with the spectral data in its most pristine state.

2006, July 14 - 09:42 — Matt Tiscareno

Re: Interpolations

Oh, you have more discussion here.