After thinking about it for a while, I've changed my numbering system because I decided the up/down nomenclature would get confusing. Here's the summary again of the status on sol 513, with a map of the places I spotted holes. The map should be regarded as "notional" -- it's very sketchy.





I started looking at all the wheel imaging campaigns that have been performed since sol 463. There are tons. Here is 463:



There has been change in that time. I don't know if the map is the way to go for looking at change. A different way of going about looking for change is to construct a spreadsheet, where I note which inter-cleat spaces were seen on which sols, and note whether they contain holes. This approach has helped me find change. For one example, there is a biggish hole at the medial side of spot 9 on the right middle wheel that first appeared between this image from sol 463 and this one from sol 469 in which the RM wheel is entertainingly perched on a protruding rock. That new hole is on the edge of visibility in sol 469; it's easier to see in this image from sol 476. Incidentally, this is the same one that they targeted for RMI imaging on sol 502. The right middle wheel is easy to track because it's the easiest one to see from the mast-mounted cameras. In a sense, it's fortunate that the rear wheels appear to be less damaged, because they're the hardest ones to see.

I'm persuing your approach of the pdf template.
Here a 720 degrees version of this Sol 513 MAHLI image:


Not yet quite perfect. But composing/stitching four appropriate fragments of this kind of images could return a geometrically almost normalized view of a wheel mantle, hence allowing a better comparison of changes, or even providing the basis for a texture applicable to a 3d-model.

Pretty cool! That twist in the middle was quite a puzzle, until I realized you're projecting the inside surface of the wheel in the part of the view where that's what's visible, and the outside surface where we see that, and the brain-bending twist is where the viewer's sightline is tangential to the wheel's surface. Nice. It looks really excellent in the positions where it's at a moderately high angle to the camera.

Thanks for explaining the image! I could't have done it better, at least late in the night after a long day.

The underlying idea of the projection is, to wrap a rectangle around a cylinder, bend and translate this cylinder a bit, until it matches with a wheel in 3d relative to the camera, then simulate the mapping of the wheel by the camera. To put the color of a pixel within the rectangle, get the color of the raw image at the described transformed position.

My simulation of the wheel and the mapping by the camera needs some more refinement to get a well-standardized geometry.
I'll work on this in small steps, and let you know, when it's more or less satisfying to my eyes.

I am sorry, before posting I edited my reply into obscurity.
I mail that part over to you.

@elakdawalla: Thank you and I prefer the new number system.

Emily, this "map" of the weels is so good! Can I use it in a space exploration french forum to illustrate theses weels anomalies? Citing that this is your work, obsviously!

Next step of Sol 513 MAHLI left center wheel image processing:

By wrapping the wheel into a cylinder-like surface (radius varying according to an appropriate parabola), and projecting it into the raw MAHLI images

the surface of the wheels gets geometrically standardized to a near-rectangle


Cropping and stitiching results in a transformation to a standardized geometry (similar, but not quite identical to the cylindrical map projection):

Seriously cool - and exactly what I envisioned a wizard could do!

Now the next step is obvious - repeat for later sols and animate to see the changes. Hopefully you can turn at least most of the steps into an automatic pipeline so it's not a tremendous amount of work for each set of images...

Thanks!
The manually difficult unwrapping is mostly automized, reduced to a couple of parameters which I need to adjust between images. So I'm optimistic, that the steps you're suggesting become feasible.

Thanks, very nicely done. Hhope adapting the process to the remaining five wheels won't be too much further effort.

I didn't yet test the other five wheels. It's probably one or two further parameters to be adjusted, mainly the distance to match size and perspective.
I'm dreaming of a fully automated feature recognition subroutine to eliminate any manual adjustment, but that's certainly more than a few hours of work at a week-end.

https://xkcd.com/1319/

i c u r doing this job since quite while.

I have a few questions about the wheel design on the current rover and the next rover:

1. Was there a design spec on having to navigate dunes? Is so what was it?

2. Has any thought been given to modifying the next rovers wheel design to make traversing a dune "easy"

3. Has any thought been given to use of "tank tracks" instead of individual wheels?

Actually point 3 has been already discussed on the MER forum.

The problem with "tank tread" designs are the mass penalty and, more importantly, treads can get thrown or break.

When that happens then your rover becomes crippled. With the current design, even if one wheel fails you can still move.
Spirit was able to do this for years