Updated route map (sol 875)

Click to view attachment

Er, Tesheiner, did you check Ft Gibson size from the Navcam images yet?

My parallax measurments are:

to furtherest outcrop exposure north of Oppy is 32.4 m +/- 2.0;

to furtherest outcrop exposure south of Oppy 6.3 +/- 0.1 .

I chose to use the northern-most visible block as the start of the feature, reasoning that from orbit even a few patchy blocks will show up very bright.

so the outcrop is 38.7 +/- 2.1 meters long.

Looking at the feature I think the faint white area to the south of your sol 869 position is the southern end of the feature.

More troubling are the parallax measurments to the nearest outcrop from the current ( sol 875 ) position:

I measure 14.5 +/- 0.1 meters to the nearest outcrop and 25.8 +/- 0.3 to the other side -- for a width of 11.3 meters. There is a fairly large dune ripple down the center.

I think that Oppy is about 13 meters further south and a few meters west from your current position.

I saw the north looking images but didn't have time to double-check that sol's position; I'll do that tomorrow.
Those outcrops are good opportunities to sync' the map.

QUOTE (Tesheiner @ Jul 2 2006, 09:45 PM)

I would try a similar aproach as for Spirit.
You must find similar features (e.g. rocks) on both pre and post-drive images and calculate the net driving distance using any Parallax Calculator. For the driving heading I would take the beacon as a reference.
This approach, together with periodic corrections (if possible), worked quite well for Spirit's route map.

If the apron is as smooth as "we" suppose it is, "they"'ll probably head very stait to a dirrection and keep it till VC. So when we'll know the heading, counting weels turns/marks will also help knowing where we are.

QUOTE (Tesheiner @ Jul 2 2006, 09:45 PM)

...you'll be back for VC arrival on Sol 935.

The problem is that rear-looking navcams are not always taken and take quite a long time to be downlinked; and using hazcam pics to count wheel marks is only doable for short drives.

Who cares? The apron will after all be crossed in two days
(Or it could be)

Updated route map (sol 876)

Click to view attachment

Horton, I did some re-adjustments on positions 869-875. Have a look on them.

Definitely looks better to me.

I measured the distance to the center of the dune crest down the center of the outcrop to be about 11 meters from Oppy and then drew a line parallel to the dune center 11 meters from the center.

I then measured 19 meters to the southernmost visible end of the outcrop -- and 33 meters to the edge of the small crater to the south. the arcs from these two points intersect the 11 meter dune line at about the same point -- about 5 meters to the south-west of your position.

Of course -- as always -- this assumes I am actually seeing corresponding points on the map.

I've been looking at the route map for some time and I just can't see oppy making a beeline for Victoria's rim. Wouldn't it make more sense to "connect the dots" visiting the two small "waypoint" craters in the apron along the way to the rim?

Updated route map (sol 877)

Click to view attachment

The light! The light! They have seen the light!

I don't think the apron is remotely smooth. We can see from Tesh's photo-map and from the recent surface pan cams that it is covered in dunes, at least in the area above the white "scarp", behind (south of) Beagle Crater.

Kenny

I agree with your 877 position -- but not the 876 position.

I measured a change in position of 24 meters +/- 1 meter, using a number of before / after move pancam points.

Curiously, where you place the "kink" agrees with my 876 position...

I always do my position calculations before I look at yours and it is reassuring when they are the same -- and distressing when they are not.

Now that we are again in a "feature rich" area, there should be no more differences.

i dont think those pics are aimed directly at the apron around the crater.....

I pinpointed sol 876 position with the help of a polar projection; that's what I usually do when we are at/near outcrop rich areas.
Here you have two images of the process. The first is the polar projection and some heading lines (in yellow) to a set of reference features (cyan points); the yellow points are placed at the actual location of those features, as measured by parallax (AlgorimancerPG tool). The second image is the same thing but the polar projection's layer hidden.
To locate the referred sol's position, the trick is to move the heading lines' layer until the points fit with the background map image. The second image has the headings layer actually located at what I believe is the "best fit" position.

Click to view attachment Click to view attachment

The dune tops look very soft to me, best to take care

I noticed that the NASA updates have been very sluggish lately on Opportunity. Perhaps, it is as simple as someone being on vacation or maybe there is some uncertainty on how to best approach the big crater. Anyways, what are people's thoughts on Beagle crater, being more specific, do people think Opportunity will enter this crater? I would be very interested to read what the people on this board think.

BTW, are there any other Red Sox fans here?

That's a damn clever and efficient trick. Thanks for describing it.

Tesheiner's method is the fundamental method for locating planetary landers, and has been ever since Ewen Whitaker did it first for Surveyor 1 (and strictly speaking, by Jaffe et al. just before him, but they got a bad result). But it does depend on being able to match features between the pan and the overhead view. A lack of such matches is the reason Luna 9 couldn't be located in the same way.

Phil

Route map, updated to sol 878.

Click to view attachment

PS: This is my last update to the map 'till mid August; I'm starting my summer vacations.

Thanks for all the map updates, Tesheiner! They've been an integral part of the long journey south for many of us.

I hope by the time you resume, we'll be at Victoria!

I second that!

Ditto! I check this thread every day to get the latest.

I go here several times everyday to look for new maps, you're doing a fantastic work!

Not that I'm in pessimistic mood but I think Tesheiner will be back before Oppy is on the rim of Victoria Crater...

WHAT!!?? Tesheiner is going on vacation? Who approved that?

this is going to be a hard time without our daily fix of your Route-Map updates
In the mean time, I for myself can't even think about the progress of the MER mission other than in
the daily rythm of your fine map updates ... actually it's the first internet address I click on every morning
too look for a new Tesheiner Route Map update

[/quote]
I hope by the time you resume, we'll be at Victoria!
[quote]

don't worry: you won't miss the Victoria approach ...
by mid August, we will probably still be in the middle of the Beagle-Crater-Campaign

Don't worry, there are plenty of people to carry on the route map work - I've covered for Tesh in the past but there are many others as well.

I'll be able to do it for a couple of weeks but then I'll also be off till mid August! (Really hoping not to miss Victoria arrival, note to rover drivers - not before August 16th, ok?)

James

Thanks Tesheiner for your dedication. Have a good vacation away from the Mars front.

James, thanks for picking up the cudgel and leading the charge. I will try to keep your back safe by continuing to hack my own route map and yelping when I think there is a difference more than a few meters.

I usually proceed by determining the direction of motion by comparing before / after panoramas.

pancam / pancam comparisons are the best, but pancam / navcam can work too.

I then determine the distance with a set of parallax computations of selected points in both before / after 3D pairs and then apply direction of motion cosine corrections.

I usually can determine the direction to a few tenth's of a degree and the distance to a few meters -- unless of course Oppy is in featurless dunes -- where the major difficulty is finding suitable features to parallax measure.

I guess you've all noted that Tesheiner went on vacation when we were in a place where we can see a lot of features so we'll not be lost! That's what I call dedication

I am back from 2 weeks of vacations. I am surprised of Oppy's route. It has continued toward the south. Not yet I have reviewed all posts to analyze what are the reasons for not taking the gate toward to BC after reaching a pool of black riddes as the shortest route.

Rodolfo

I've been looking at the source image for the current route map, found here:

http://www.msss.com/moc_gallery/r10_r15/im...5/R1500822.html

There is a map projected and raw version of the image. Down at the bottom of the page is listed derived values such as scaled image width & height (kilometers), as well as the pixel width (in meters). I'm finding that it is a bit unclear whether to apply these to the to the raw or the map projected version of the image. It would sure be nice if they applied to the map projected version, but if so do the scaled width and height apply to the actual image width and height, or to the rectangular region projected onto the image? The fact that the term "derived values" is applied to these things suggests to me that they in fact apply to the projected image as a whole (good!), but I haven't found a source elsewhere on the site that neatly confirms this.

Any thoughts?

algorimancer, looking at images which are very long and narrow, where the rectangle is much wider than the map-projected image itself, it seems that the values apply to the original image, not the map-projected version.

Phil

Okay, that being the case, I find that the projected version works out to about 0.751 meters per pixel (in both dimensions). Good to know.

This is a bit off topic, but that comment reminded me of something from Doug's last interview with Jim Bell that I thought was worth mentioning. I hadn't noticed it mentioned elsewhere in the forum, and it is probably not worth starting a thread in the tech, general, and imaging section.

Doug asked Jim to clarify the true field of view of the Pancams, since various published numbers have been seen. Jim said the key number to remember is 0.273 mRadians/pixel. With Pancam images 1024 pixels on a side, each pancam should have a fov equal to precisely 16.01715 degrees.

Since this is kind of important to know with confidence, I emailed Jim Bell as to which MOC image to apply the scaled measurements to. Here is his response; looks like Phil was correct:

"I wasn't sure of the answer myself so I checked with Tim Parker at JPL, an expert on the use of MOC images. Tim says that indeed you should apply those scales to the unprojected copy of the file, not the map-projected copy.

Hope that helps,


Jim Bell
Cornell U."

Is that field of view measured from the centers of the two extreme end pixels (total width of 1023 pixels)? Also, is there any distortion that makes the width of the outer pixels any wider or narrower than the pixels in the center?

That should be 1024 pixels times the pixel width, so the 16.01715 degrees figure would be from the left edge to the right edge (or top to bottom) of the CCD. As to distortion, for the pancams the distortion is effectively non-existent. It is more distinct with the navcams (but not too bad), and substantial for the hazcams.

As far as the fov goes (sorry for stretching this off-topic tangent) there must be some distortion with pancam, and I'd be very surprized if it were negligible at the 3 significant figure level of the quoted value 0.273 mRadians/pixel. Bell was ambiguous about this in the audio interview. Remember too that some of the calibrated images include camera lens distortion corrections - these are called "linearized".

So the questions remain: does the quoted figure refer to the centre of the fov, is it an average over the frame, and does it apply to the linearized frames or not?

http://www.unmannedspaceflight.com/index.p...ost&p=61030

The Pancam facts

Doug

Does anyone know when we'll be moving again?

Tomorosol

Still no Exploratorium.

So here is the route map based purely on the tracking data (36m S, 2.5m E) (Good guess algorimancer!)

Edit: version based on images is two posts further down...

Click to view attachment

James

I'd go along with that, James. It makes sense that they are not going to try to cross several ripples and would tend to drive along a ripple trough until they reach the evaporite exposure at the leading edge of the ejecta apron. Once we get imagery we can make the appropriate tweaks to the WAG route...

--Bill

OK, as soon as I put up that map, Exploratorium come back up.

Here is my best guess from sighting features for Sol 883.

I put Oppy about 7 meters south and 2 meters east of your position.

Measurments:

37 +/- 1 meters movement from measured parallax changes on a heading about 8 degrees to the right of the center of the crater just behind the 15 meter black dune crater ( BC) ;

49.5 +/- 1.2 to the front of the dune intersecting the crater on a line through BC;

Consistancy checks:

41.5 +/- 0.8 on a line 28.0 degrees from BC through Oppy position intersecting the left edge of the dune abutting the 15 meter black crater.

Also, visually, Oppy is now just in front of a fairly long stretch of outcrop -- the faint white line on the map.

You mean somewhere near my point based on the tracking data alone?

Problem is, the angles are all wrong down there. Everything fits nicely up here.

I too came up with a figure about six meters further south. There's room for error though, I'd feel more confident if there were some images in the direction we drove from; I'm having trouble identifying some rocks with confidence.

Hi Algorimancer.

Just wanted to thank you for your wonderful parallax tool AlgorimancerPG. I used it exclusively for the parallax measurments in my sol 883 position estimation.

I used the online tool MER Stereo Parallax Calculator for a long time, but discovered it was not very accurate for over 50 meters measurment. ( I routinely subtracted a 2 pixels correction. )

FYI, I remeasured several rocks after I refined the direction of motion and got a movement of 37.2 +/- 1 meters on a heading of 7.9 degrees right of center of the reference crater. Didn't make much difference in the original guess. The biggest error source is the center of the 10 meter reference crater.

Sol 884 looks like a 39 meter move just to the left of the 15 meter circular black dune. I thought it was a crater.

A Navcam parallax measurment to the closest rim of Beagle is 75 -15/+22 meters ( 75 meters as measured on the map )