The HRSC on Mars Express has a resolution of only 10m (2m on selected targets), not enough for our current purpose.

Could we take a "Galaxy Zoo" approach and divvy up each chunk of terrain among multiple volunteers? (Judging from the posts on this thread, I'll bet there's a heckuva lot of volunteers rarin' to go). Using Phil's template as a guide, each volunteer makes his ratings, then sends back the data. Multiple volunteers per terrain unit could allow a statistical average to be developed for each terrain piece.

(And like the "Galazy Zoo" approach, a training set could be used for training, and for validating that the volunteers are ready.)

-Mike

Not in the kind of resolution needed to actually see and characterize the ripples. For that, you need the HiRISE camera. (At least, if the MEX cameras, even the HRSC, can image something as small as the ripples in enough detail to do a stereo image that gives you the size and extent of individual ripples, I'm certainly not aware of it...)

-the other Doug

Those who suggested that clickworkers was the closest analogy were correct. Basically you just need an "impassible dune" cursor for image markup. Galaxyzoo and stardust@home are similar, but galaxyzoo doesn't use clickable images and stardust@home allows a single click on a movie.

I certainly hope that you, like Rui, decide it's funny, Stu. It was definitely intended to be.
Seriously though, you two are far and away the most poetically inspired and poetically eloquent members at UMSF, and I would love to see the outcome of a period of collaboration between you, in giving expression to the spirit that pervades this forum. The desert island is only one idea for a venue, and comes to me because I can see out my window a perfect example in Mokoli'i Click to view attachmentfloating serenely in Kaneohe Bay. There is exactly one coconut tree on the island, in case you need to settle any 'artistic differences' at ten paces.

Unsharp masking could be useful for the chunks:

GIF format

Don't worry, just kidding, I thought it was really funny.

Not as funny as Rui's follow-up sketch... that had me chuckling all day during a rather challenging and not-exactly-laugh-a-minute Course about constipation and caring for terminally ill patients (and you lot all thought my life was full of poetry and martian daydreaming, eh?)... god, we'd be like a space geek version of "Withnail and I"...!!!

Bundling up comments on several posts plus my own ramblings; apologies for the length.
It seems support is emerging for two approaches:
1. software automation

Pros:
• Can probably be made to work effectively;
• Spits out normalised data set covering as large an area as you like in two minutes flat. (To a first order of hand-waving anyway )
• Easily testable, reproducible, and improves over time as the author tweaks the algorithms

Cons:
• dev lead time. It may take some time before the first useful two-minute run!
• only a few people here have the code-fu to do it, a single-point-of-failure. FSM forbid James falls under the proverbial bus just before the first release in a month's time, but...
• lots of possible algorithms that could be combined in various ways, with various parameters and feedback loops, make it easy to try lots of variations and select the best.

2. A Mechanical Turk, human-driven distributed process. (several variants on this have been suggested.)

Pros:
• enables lots of people to contribute
• very quick and simple to kick off (a completely manual version could start in 10 minutes time.)
• the raw data could be straightforwardly fed into software (from 10 lines of Perl, to full-blown C++ GUI apps) for further processing and display in various styles.

Cons:
• indeterminate runtime. Might take weeks, or (pessimistic assumptions) five people, managing one grid square per week, works out to roughly 25 years! (OTOH,.. we could prioritise the closest areas of terrain - a bit like a chess algorithm, in fact, selecting and pruning a tree of possible future moves... )
• Once certain parameters are set (e.g., grid square size, the particular archetypal images are used for the classification cheat-sheet, etc) they can't be altered without junking whatever data set's accumulated to date.
• People make mistakes and/or differ in interpretations. (A training stage would take time to build.)

I suggest that we do both. There's no reason we couldn't start on a Mechanical Turk process whilst the bit-twiddlers do their thing; at some point (a week, two, ten,..) the software's results become more useful than the manual ones, and the baton is passed. A purely manual process could start whilst work is underway on a web-based, semi-automated, galaxy-zoo type solution, AND a dedicated program to do the whole thing.

I hate it when people say this in email at work, but as I've already used the word "solution"... ... "thoughts?"



Correct me if I'm wrong, but my impression is that all the dunes encountered on the way to Victoria had the same angle of repose, so the visible width has a linear relationship to cross-a-bility.


$wine{'cava'}->spray("nose");




Question: which part of Endeavour? I've been assuming your aim point is the jagged western rim peaks; is that right? Is there a definite aim point, or a defined area to aim at?

Here's my rudimentary attempt at showing a proposed track. For more detail at the "points" of interest, zoom in on the HiRISE image at:
http://marsoweb.nas.nasa.gov/HiRISE/hirise...SP_001414_1780/

At this point I don't seem much merit to tracking beyond the giraffa camelopardalis, except in very general terms. ...lots of options.

Thank you, Paolo, for clearly stating your wishes, and thank you Oersted, Imipak, Tesh, Phil and the others who have made clear, practical contributions to fulfilling those wishes. An area map of favorable and unfavorable terrain for Oppy (especially if it can be generated by automated software) will be a major contribution to planning the course of rovers in this and other missions. I have no expertise in this higher order image processing, but I would hope that a way could be found to include points of scientific interest - e.g. craters and cobble clumps - as map features so that a planned track can incorporate them, if the PIs judge them to be worth closeup examination. It is probably too complex a task to automate selecting points of interest, but I would be willing to manually circle any such features I could resolve in a HiRISE mosaic. Points that are isolated by impassable ripple zones could be easily bypassed in preference to those lying in 'safe' zones.

"Some sort of Fourier analysis" on subsections of the HiRISE has been mentioned by Paolo and James. I think that's a good approach but I can't visualize how you would define the parameters for the transform in order to get algorithm for the sort of texture matching that's needed. The general idea though seems to me to be a good approach to try for any of us who have some coding ability. If it works we can then process the whole thing repeatedly (and at various scales) and get something that can then be thrown at the full traverse HiRISE images.

Geert's comment in post 168 got me thinking that something like a Hough transform would be a better way of identifying the combination of scale, orientation and consistency of each subregion. I did some work with Hough Transforms last year and they're simple enough to implement so I'm thinking that I'll give that a try but I'm absolutely swamped with work at the moment ( I'm working with Banks these days, which is interesting to say the least) and I'm unlikely to be able to build anything competent in the timescale we're looking at so I'm just throwing this out there in case anyone else is in a position to experiment.

I've been working for the past two days on trying to get the JP2 into Python (Google App Engine) so help with installing the JasPer plugin for ImageMagick would be greatly appreciated. ImageMagick can automatically split an image into hundreds of named tiles, but it can't read JP2s without JasPer. And JasPer is only available as C source. The JP2 format is evidently quite a processing bottleneck.

Anyway, I can start development on the mechanical turkish front-end without the images (does http://helpoppy.appspot.com sound good?); I just don't want to be redundant, so if anyone else is attempting the same thing please speak up.

Hmmm- it seems like Fourier analysis would handle this problem rather well- the dunes should provide a strong peak in the Fourier transform, and the wavelength and amplitude of that peak, if it could be identified automatically, should describe the size and density of the dunes (which are the questions of most interest for driveability) quite well. That's the theory, anyway...

John

For the laypeople among us (that's me) - is that sort of pattern essentially why dunes disappear when I've zoomed out to a certain point.

So - 100% - the flat parts look flat, everything else looks lumpy
Zoom out a bit - maybe 25% - the smaller, navigable dunes vanish, but the bigger scarier dunes are still obvious
Zoom out to 6.25% and it's almost all smooth

Doug

Yes, that is what I was thinking.

I've started doing Fourier transforms on 256 pixel (east-west perpendicular to the ripples) slices and have confirmed to myself that I can at least detect ripples. e.g.

No ripples:
Click to view attachment
Most of the power is on large scales

Ripples:
Click to view attachment
Big peaks at 10-15 pixels (the scale of the ripples)

I'm sure the signature will change significantly as the ripple size changes. It's bedtime for me now, but tomorrow I'll put the code in place to run these FT's over the whole image and then make maps of various characteristics of the Fourier profiles generated.

James

Wow. Just amazing James.

Have you wondered if maybe some poor guy on the MER team will be laid off once you complete this?

When James also gets a job with Google, we'll just have to expect him to use his company-funded free time on just this sort of thing.

Darn. I wish I'd paid more attention in math class...

Is there any 'brute force/raw time-consuming' stuff I could do to help? I lack programming skills or any knowledge of Fourier transforms, but I have free time...

As I have little to no knowledge of Fourier transforms, I made a much simpler algorithm and applied it to a screenshot of the IAS viewer. It is hard to believe that it would be of any use, though, as it seems easily confused by the bright spots.

Click to view attachment
Click to view attachment

Wow! This is amazing.

I was myself just trying to code something like this, running a Fourier or a Hough transform on the HiRISE images, I once tried to do something similar to detect ocean waves from satellite imagery, but I'm nowhere near yet to producing anything like this, it's a lot of work and like anyone else I'm swamped with other jobs.

But I guess what we might try next is running the same analysis on a 'known' traject, let's say part of the road to victoria, then we can see how such an analysis corresponds to known terrain, making certain we are on the right track.

I swore to myself that I wouldn't mess with the route map proposed in post #209, but after running a little pattern recognition program to the SE of Sophi, I noticed a feature that was missed earlier. ...wondering if a jog to the cluster 260 m SE of Sophi is warranted, if a more easterly route to the South is even feasible.

I've attached version 2.0 of my proposed route, with the hope of seeing some of the experts here give this a closer look.

The drive from the second to third points of interest would certainly be a "nasty" piece of work. There are some big dunes here, but also enough pavement to make it appear that there's actually several plausible tracks through this tight spot in the maze.

I don't think this has been mentioned yet, but I gather we are interested in quite broad characterization here, eg "easily drivable" vs "purgatory-ish". Fiddling with imagej, I noticed that the smaller ripples (which tend to have shorter crest to crest separation) can be quite easily eliminated from an image by gaussian blurring with kernal size a bit larger than the separation. The bigger, nastier ripples are still visible after this blurring. Perhaps this would help either manual or automatic approaches. After the blurring you could use some tool such as FFT or a tool that measures the variance across the image to find the areas that have most "power". I expect this to give you problems around exposed bedrock, though.

I do not want to impose a "mission statement" . I'm a guest here, and am trying, like you, to do something useful for this long traverse. In the past having such a traversability map has been very useful and relatively painless to generate and am sure this time it will be more difficult. It worked pretty well in the past but I do not want to suppress random ideas. I have to say that suggestions for science targets are way above my head, I do not have the qualifications to understand and evaluate those. I'm pretty sure some scientists are following the forum and they might pick up some ideas.

Regarding comparing drives pre-VC to drives after VC, I think this is best done by us drivers. We have learned so much after we loaded the new version of the FSW in 2006 which we have been using since sol B950 and we also have learned new tricks. Unfortunately these are details that very likely I cannot discuss in public an even if I could it would require extensive driver training to appreciate. I do not mean to be dismissive, no greater pleasure to explain all the technical details to Mars geeks, it is just rules I have to follow to keep my job.

I also have to say that unfortunately poetry is completely lost on me since English is not my primary language (I'm Italian).

Paolo

I'm all for the Fourier transforms and such, but I also think it's important to check these areas visually, as well. And for one good reason -- almost any mathematical analysis is no more reliable than the base data it has to work with.

I have in mind the following scenario: We run Fourier analyses transverse to the predominant ripple orientation, since that's what gives you crest-to-crest distances and also height estimations. But it also fails to capture, in any meaningful way, the same type of information for the crossover ripples that close off all the inter-ripple lanes after only a few tens of meters.

Remember how we progressed through the Etched Terrain? The inter-ripple troughs would run for 10 to 20 meters, and then a crossover ripple would close off that lane. We had to then climb over a ripple to get to the next best trough over, one side or the other, if the crossover ripple appeared impassable (and it often did). When I look carefully at the full res images of some spots along the south route, I see the same kind of crossover ripple activity, and thus we'll see the same need to climb these ripples if we want to continue to move.

But because of their deviation from the general north-south orientation, I'm concerned that perhaps Fourier analyses aren't likely to collect enough data about these ripples to be informative of the real trafficability. That's why I think we also need eyes looking through these mazes, to find likely routes that involve climbing over ripples we're certain Oppy can handle.

But, on that score -- if, as Paolo has said, we have to keep in mind the 5-wheel scenario, we're really going to need to know just how easily a 5-wheeled Oppy is going to be able to climb or cross ripples. Without some feel for the real impact dragging a wheel will have on crossing ripples, I don't know how easily we can project a safe path into heavier ripple fields.

I will say this -- I've taken another good full-res look at the terrain to the east-northeast, and while it looks very smooth at lower resolutions, when you look at it in full res you can see it's covered with ripples to at *least* the degree we saw at Viking and Vostok, and they, too, have a predominant north-south trend. If we can't handle moving steadily transverse to ripples of that size, then I guess the south route may be the best of a not-wonderful set of choices after all.

Actually, one of the bigger reasons I had for wanting to take the east-northeast route above the worst of the ripple fields is that there is actually some exposed outcrop on rim remnants at the very northern extent of Endeavour's rimwall. What we see along the northwest quadrant looks a lot like a rim landform eroded way down and then covered over by the same evaporite groundcover that we've been seeing all along. Just seems to me that it's overall less mileage to more ancient rimwall outcrops if you go by the east-northeast route, since you won't have to backtrack back 5 km north to get to them in the northern rimwall, or go yet another 10 km south to get to outcrops on the southern rimwall... Of course, HiRISE images of the various sections of the rimwalls will help us decide just what we're actually targeting first at Endeavour, so perhaps we should wait for those before making the final decision as to the direction from which to approach.

-the other Doug

I'm fine with solutions as long as Oppy is not part of the precipitate.



That's my recollection as well but geologists that I have talked to said I should not count on this. My hope was that since the slopes on the sides of the ripples were pretty much the same, the spacing between the ripple crests was an indication of their height and ultimately their traversability in the east-west direction. Alas I have been told there is no scientific foundation for this. At least tho, spacing of ripples will tell me how ifficult it will be to make it between ripple crests.



Very good question. If, and that is a big if, we can make it in two Earth years, we would get there by martian wintertime and therefore northerly facing slopes would be preferred. In that case the south rim would provide plenty of northerly facing slopes. Now that I think of it, I am assuming we would head straight into the crater at the large NW facing opening in Endeavour. My assumption might be completely wrong and probably there are no clearly defined ideas. Sorry! All this is under construction. More questions than answers! I will investigate and let you know.

Paolo

That is correct. I would use an automatically generated map as a starting point, not as a goal. Visual inspection will be a major part into this.


You are absolutely 100% correct. Unfortunately the HiRISE imagery does not provide enough details to tell us if and where we can cross over between ripples. The DEMs (digital elevation maps) I can get to do not have enough resolution to indicate possible passages. Only PANCAM and NAVCAM will tell us that. We can only guess that a field of tall ripples is less likely to provide passages between ripples, while smaller ripples might provide some.

While the stock market always warns us that past performance is not guarantee of future returns, we have no choice but to hope that Mars is less capricious and we will find some passages when we are there. This is why I ws saying that trying to define the exact details of the path is very difficult and prone to being revisited on a sol by sol basis.



Hmmm except the experience with Spirit, we do not know how difficult is to cross a ripple with 5 wheels. We never tried and when we did testing in our sandbox for Purgatory we quickly found out that it is not trivial to do this types of testing (our sandbox is not yet equipped with a gravity field modifier). Once we are back in ripple field we might test 5 wheel driving on Mars, but I think it would be difficult for me to sell to the project.



You are again correct. The major difference between the drive from Endurance to VC vs the drive to Endeavour is that the former was north-south while the latter has a significant component in the east direction. When ripples are involved, this spells trouble. My hope is that with a map like we are talking about, we can find sections that would allow the rover to move east as much as we can.



I agree. I think I said it earlier, and if I didn't I'm sorry. I think it is best if we prepare our tools, think up a procedure to analyze all the HiRISE images we need. In the meantime we can test our methods and procedures on the current HiRISE near VC. We can compare our predictions with actual drives and make appropriate changes to our procedures. Once we get down the new HiRISE we will have tools and procedures somewhat refined.

Paolo

I'm sorry I can only contribute a pair of eyes.

I looked at the S-SW side, with IAS Viewer at full resolution, just following the smallest dunes:

Click to view attachment

The ring of exposed rock around Victoria seems to have a mix of big and small dunes. The ring is narrower at roughly the same point where a path would cross it aligned N-S with the dunes.

The map doesn't look so big to check manually if broad areas can be discarded because they are enclosed by big dunes.

I realized that I might have potentially alienated some people when talking about automated process, FFT and the like. It was not my intention. I was thinking aloud while typing. So here is in lay terms what I was saying. If you look at specific areas where ripples are, you will see that there are ripples that are closely spaced, some are farther apart, some have clearly visible crests, some do not. But to your eyes they look different, do they? There are image processing and machine vision algorithms that can mimic this capability. FFT, wavelets, Gabor filters... They are not as god as the human eye, but they should be able to see the differences here. Writing a piece of code that does this can be pretty complex. Fortunately there are public domain packages that should be able to do this: ImageJ, the Intel Image Processing Library to name a couple.
Matlab ouf course can do that as well (but that is expensive!).

IF we can do this, it would be great. If t gets too complicated, not enough time or resources are limited, I understand. It is a big problem even for people like me that spend their day dealing with it.

Paolo

Wandering through the Hires image using the IAS viewer the height of the ripples appears to scale with the brightness of the east-west waves visible at 1:8. Where the waves are visible the ripples look like they did near purgatory or higher, areas without waves look more like the area around vostok. The waves also look like they are higher than their surrounding. If I remember correctly Oppy got began digging in at purgatory when she started climbing one of these waves. If I was the rover driver I would be thinking about avoiding as much of the wavy areas as possible by looking for a path southeast through the etched terrain like in Tesheiner' post from a couple of days ago http://www.unmannedspaceflight.com/index.p...st&p=126410

I also think that Fourier transforms can help. It occurred to me that we may not be exploring new ground here. Sorry to sound like the unwelcomed guest, but has anyone considered that NASA/JPL/DOD/etc has already, long ago employed various scientists, engineers, graduate and undergraduate students to explore the concept of automatic terrain recognition? Surely the remote sensing industry already has software that can do what we have been trying to do.

Not me I'm afraid. I don't have the coding skills for that (and I've fallen at the first hurdle trying to find a good way of slicing up these jp2s).


Re: automation
The one issue I have with Fourier Fast transforms and other automated processes is that they may show different responses to ripple formations over bedrock compared to sand. Even if the ripples are morphologically similar, the increased contrast between the dark drifts and light bedrock will produce a stronger peak (albeit at the same wavelength).

I am convinced eyes (human as well as oppy's) are absolutely necessary to find a route through these mazes. However, it's a 'large' area and there is an infinite amount of route-choices, and mathematics can hopefully at least aid in lowering down the number of options. The 'mathematics' are just one method to try, while others will use their eyes, basically I think any attempt is worthwhile just to see where it takes us, and, in a way, it's just 'fun' to do.

It looks like a Fourier analysis can tell us something of the 'wavelength' of these ripples, as I hoped it would, that's transverse to the predominant ripple orientation so it won't tell you how often you will encounter crossover ripples. However, if these ripples behave anything like 'normal' waves, the distribution in the Fourier-spectrum will probably tell you something about the amount of crossover ripples. In other words, in an ideal situation if you find only one 'ripple-frequency' in the Fourier spectrum, all ripples will have the same wavelength and you can expect long open traverses in between the ripples. In a bad situation, when you find a very wide frequency-range in the Fourier, you can expect a far more chaotic terrain with few open stretches and a lot of crossover ripples.

Once again, supposing ripples behave anything like sea-waves (or, more specifically, waves in shallow water), there is indeed no clear mathematical method to relate their height to their wavelength (distance from crest to crest), the Fourier analysis will tell you something about the wavelength but it doesn't say anything about their height. You expect 'long' ripples to be also high but mathematically that doesn't necessarily have to be so.

I proposed 'testing' the analysis on a part of the route to Victoria simply to see whether we are on the right track, I fully understand that Paolo doesn't have the authority to share driving-experience with us, but if we run an analysis on the Etched Terrain and our analysis tells us this is a wonderful area for driving, then we are way off and should do a lot more number-crunching before we can use it on the terrain ahead... Also we might be able to use an analysis of the Etched Terrain as a benchmark to compare the terrain ahead to.

You're a wizard! Click to view attachment
Perhaps it's in your plans but if not, may I suggest to combine/average the results of 256 slices (north-south) to see what happens?



Sure a visual check will be required but only in those areas with mid-big ripple sizes.
Edit: Oops, I saw that Paolo already gave a proper answer to this topic.

Terrain classification for autonomous rover driving has been done. Several times. JPL, CMU... The DARPA Grand Challenge is a good example of this. There are much more sophisticated examples out there. Remote sensing people have the additional advantage of multispectral imaging, here we do not. But as fas as I know this has never been applied to Km scale planetary exploration.

Paolo

I wonder if a suitable North facing slope for Oppy to spend next Winter would be at the Northern end of the range of hills to the West of Endeavour crater. Oppy could then spend next winter climbing up part of this hill much as Spirit climbed Husband Hill throughout its first Martian Winter. I guess that this hill might be three times the height of Husband Hill and so it might not be feasible to reach the top. The view however over Endeavour from the top of this hill would be spectacular.

Hey, ImageMagick is one image manipulation tool I can use... Forgive a very obvious question, I've just never needed to know this until now - could someone post a link to the JP2 file(s) referred to? Are these / is this the canonical source image(s)? Thanks!

Yeah, what I didn't mention is the input for those plots was an average of 5 pixels N-S not just a 1x256 E-W slice.

I'll be experimenting later, but my current thinking is to run this FT on 8x256 sections then take some average of something (waves arms) over 32 slices to get "characteristics" for a 256x256 pixel square. Will probably also need to factor in the pixel brightness statistics which should help find areas with exposed bedrock.

Then I just need to get it to pick out and mosaic areas with similar "characteristics" and we can see what works best. (or if it works at all!)

Oh and to pick up on what others have said - I agree, running this it on the Endurance -> Victoria terrain is an essential test.

James

Here're the homepages of some HiRISE shots centered at VC. The links to the JPEG and JP2 products are at the right.
http://hirise.lpl.arizona.edu/TRA_000873_1780
http://hirise.lpl.arizona.edu/PSP_004289_1780
http://hirise.lpl.arizona.edu/PSP_005423_1780
http://hirise.lpl.arizona.edu/PSP_009141_1780

Warning! The JP2s are huge!

Dear All,

Let me give you my feeling. I understand that for the coming month or so, we have to concentrate to find out the general best possible road as defined by Paolo. This is a very challenging topic and I agree that a kind of automation has to be found out. The possibilities are too wide to try to spread the work between UMSFers eyes. Our group contribution here will be (mainly) limited to the one that get, time, talent and knowledge.

So, I’d like to encourage the “others” like me that can help with eyes, hands, brute force, etc to get prepared for the next step: we all will contribute but only when the general route will have been chosen.

The idea is to have a group of people dedicated to “explore” the intended path, say a 100 meters wide path, another group to explore the left side of the path (another 100 meters) while the third will explore 100 meters on the right of the path
When I write explore, I mean “try to get the most information of”.
The main idea will be to be about one or two weeks ahead of Oppy, not more, because, at anytime we could need a path change i.e. the information from people looking right or left which right or left will be modified if the main path changes. So, I guess that a 300 meters wide path 1-2 weeks in advance is more than enough.
People could switch groups regarding availability, interest etc, could participate in 1,2 or the 3 groups but the minimum will be to have enough people on the main path at anytime.
On top of this we’ll also need a management of the whole picture. I mean, as we’ll have data from the ground (navcam and pancam) that will show what the reality is like, that will sometimes “see” 1 or 2 weeks ahead we’ll have to integrate those data in the groups.
Let me tell you that if we can commit to such a “work”, if we can provide the Rover Drivers with consistent and reliable data, we will have a big contribution to Mars Exploration and this is very exciting to say the less.
I hope you’ll like the idea.

"at anytime we could need a path change"

Which is why we need to try and get some sort of survey done before we leave the Victoria Annulus

I wonder if we'll get to see the Explorer crater on the south side of Victoria? It would be interesting to see the zone between the two craters.

Is it just me, or do the camera lenses look a little cleaner now that we are out of Victoria (perhaps a wind gust help clean them)?

Definitively
My proposition is for during the "real" trip.

Wow, that's really "back seat driving" to an extreme!

... I have a mental image of a big bus full with people (about 40 or 50) screaming to the driver: "left on the next cross; no, no, straight ahead; no, stop! ..."

That's more or less what happens in reality when there's more than two drivers in the room in a tactical shift.

Paolo

Here's the results of my fiddling with ImageJ. I started with the hirise crop from jekbradbury's post (not sure if it's full res). I resized it to 50%:
Click to view attachment
Then applied a 5 pix gaussian blur, which reduces the visibility of the smaller, traversible (?) ripples relative to the bigger ones:
Click to view attachment
Finally I applied the "variance" filter with 40 pix width, followed by a 20 pix gaussian blur and a gamma tweak:
Click to view attachment
In the final plot, dark areas have small, traversible ripples, and middle-bright areas have large ripples. The brightest areas have bedrock.

This was very quick and dirty, and unfortunately I don't have time to do much more. There's no directional info here like you'd get from FT: I've just evaluated the total variance. It's not clear if bedrock can be distinguished reliably, but I'd think it would be quite easy to measure the max brightness (after some smoothing) in a small window and say you've got bedrock if the max is above some cutoff. Discriminating dangerous from safe ripples on bedrock may be hard to do automatically.

Anyway, this would have to be tested extensively of course...

Interesting results, Fredk!
Here's what I get when applying the same process to the image I'm using on the route map.
Click to view attachment Click to view attachment
(2.5m/pix)

Thanks for that, Tesheiner. And now we already have one important conclusion: whatever route we take to Endeavour, we should not go via the interior of Victoria!

Seriously, I have one more comment about the route (this is for after our drivability maps are ready). It might make sense to drive as much eastwards as we can early in the drive. I mean, we know we have to make significant west-east movement, across the ripples. So perhaps we should take any opportunity we can to make progress eastwards, even if it may not look like the best direction to start (as long as we won't get completely bogged down). Then later, if the ripples worsen, we'll have already made progress east and could head south instead. Also, if a wheel goes it may become much harder to move east.

Basically, the philosophy is we should move east when at all possible, and then head south when we can't move east due to big ripples.

I have a few Questions:

1. Is the area south of Victoria lower in Elevation then Victoria

2. Is the top of Endeavour lower in Elevation then the bottom of Victoria Crater???

3. If the area is lower will we discover different bedrock or layers then we saw inside Victoria

4. Where has Pando Vanished to??? last post in June? Hope he's ok?