Reprojection works better than it has any right to

[fivemack]

Y'know, if testing an algorithm which turns out to rely absolutely critically on precise to-the-pixel identification of surface features between images, maybe I should have chosen something other than Titan, an object all of whose surface features have critically fuzzy edges.

To obtain this image (click on it! it grows bigger!), I took the four photos that I photo-mosaiced incompetently yesterday, then solved for the best-fit circle to the ones showing limbs, then solved for the circle parameters for the ones not showing limbs by requiring the features to map to the same places as features whose positions I knew by reference to the limbs. I've then cropped off the bits that were black, or so close to a pole as to be absurdly fuzzy. I'm impressed at the continuity of scale and of colour.

The scale on this map is slightly too small to see the Mysterious Scratch without the eye of faith, but it's visible at the terminator on the big image.



This is a cylindrical projection, nothing like as good even as Mercator's, and with arbitrarily chosen equator and poles.

It is an odd feeling to be producing what I think may be the first maps of unexplored parts of a new world. I wonder if anyone will bother to name these craters. There are enough hundreds of them to drain the name-stocks of the Aeneid many times over.

Comments

[shimgray.livejournal.com]

It's also possible it's on the back-burner; the next Dione flyby is in October (a close one, though - 500km!), and they may be wanting to keep this back until they have some detail imagery to attatch it to, help make sense of the broader picture. (If this is what comes out of a flyby at 80,000km, when they've their hands full with Titan two days earlier, a dedicated approach/close-pass will be wonderful)