NASA Images Old Mars Landers

Iron Sun writes "Scientists have used the Mars Global Surveyor orbiter to capture images of what they think are the Viking 1 and Mars Pathfinder craft sitting on the surface of Mars. I'll have to take their word they are the indistinct blobs in question. The probes were supposedly just below the resolution of Surveyor, but they used a new trick developed last year to squeeze more detail out of the camera. The next target will, of course, be the Spirit rover."

Comments

Images of beagle

[nocomment]

And of course here's the images of Beagle for those who missed it.

Re:Images of beagle

[Anonymous Coward]

Good thing it landed in that smoking crater.

Re:Images of beagle

[deglr6328]

I feel bad for the scientists working on Beagle 2, but this sure is fun.

They can show us this

[AtariAmarok]

They can show us this, but they still can't show us the fillagried balconies on all the ornate buildings of Cydonia, or even a more detailed image of the Face of Elvis!. Come on, where are you priorities, NASA?

Beagle?

[palndrumm]

They actually know where the Spirit lander is... Surely at this point in time it'd be more useful to try and look for Beagle? Or is NASA not going to do that because it's not one of their missions...?

Re:Beagle?

[palndrumm]

Of course, if I'd actually taken the time to RTFA I would've seen the bit that said:

The procedure is unlikely to help in finding the European Space Agency's lost Beagle 2 lander or the similarly doomed Mars Polar Lander from 1999, the researchers said.

"It would be extremely difficult to find a lander for which the location is uncertain," Malin scientists said in a statement.

Re:Beagle?

[nocomment]

Well, since you can't read articles apparently, (hey it wouldn't be slashdot if not for you ;-) NASA say's this technology only works, if you know where it is already. You have to know the landscape and what it looked like _BEFORE_ you can try to figure out which dot is beagle. For example they say they know where the rover is because they parked it next to a rock they named "Yogi". Well they know about where Yogi was, and where all most of the dots were before it landed, therefore they look for the new dots near objoects where they expect it to be. Not even the ESA knows where beagle is, you have to try to pick a dot out of something that _might_ have tumbled down a crater? Fat chance.

Images of beagle 2

[Anonymous Coward]

.
.
.
. .

.

Weird...

[Henry V .009]

If those blobs at the end of the arrows are our probes, whose probes are all the rest of the blobs?

Re:Weird...

[stefanlasiewski]

Those aren't probes, those are martians running around with a bunch of teeny pieces from the probe-formally-known-as-Beagle_2 .

If you look in the upper right hand corner of the photo, you can even see a Martian mooning us...

Re:Weird...

[superyooser]

... and he is not happy.

RTFA

[PurpleFloyd]

If you read the article, it says that the technique is only useful when the object's position is already well known. After all, it's just a little dot. Even assuming the Beagle didn't break up completely into pieces smaller than the resolution of the camera (even with the new technique), it would be pretty much impossible to find; the area to be searched is basically the whole of Mars. I'm sure that NASA would look for Beagle, were it possible to use this technique to do so.

Re:RTFA

[mOoZik]

You're wrong, actually. The parachutes on the Beagle will be used to find it, as they were bigger than the resolution of the Mars Express camera. Furthermore, Colin Pillinger suggested that they can even use the gas signatures of the retrorockets to identify the locations.

P.S. Pink Floyd + Deep Purple rule. :)

Re:RTFA

[Anonymous Coward]

The parachutes could have failed to open. It is hard to know without detailed study of what may have gone wrong.

The claims in news stories saying that it probably landed in a particularly bad location are wishful thinking on the part of the designers. While it is a likely explanation for what happened, it is certainly not the only possibility.

Of course we should still look for it, to try to find out what happened... but we may not be able to find it due to whatever happened to it.

It's a great stretch

[dtfinch]

To argue that those 2 pixels are the Mars Pathfinder, with all that random noise in the photograph.

Re:It's a great stretch

[ColaMan]

well.... if you imaged the same spot twice on two different orbits and the pixels in question were still there, you could then argue that they are surface features.

Re:It's a great stretch

[Valdrax]

There are actually several techniques to filter noise without losing significant spikes. One which I've worked with before is the minimum noise fraction transformation, which is a modified version of the principal components transformation that works based on maximizing signal rather than variance.

"Basically," you create n-dimensional coordinate system where n is the number of colors per pixel in your image and each axis is numbered by the brightness of that color in each pixel. You plot each pixel as a point in that space. Then you take your point cluster, and -- here's where the nasty magical math handwaving appears -- you rotate it so that the the first axis becomes the line through the data with the highest signal / least noise, rotate it so that the second axis becomes the next orthogonal line through the date to have the least noise, and so on.

(A principal components transformation does the same thing with highest variance instead least noise, and the most common implementation of an MNF transformation is actually a noise-adjusted PC transform. Basically, you reorder the data so that the highest signal becomes the highest variance.)

You then take the resulting image which has the highest signal in its first layer (the first "color" value per pixel) and the highest noise in the last layer and you filter out the noise layers (determined by looking at the eigenvalues for the image). You can either average them out, zero them out (inaccurate, but a useful compression mechanism), or you can run a smoothing filter over them. When you rotate the image's point cloud back into its original orientation, you get an image which has had a significant amount of noise removed.

Once you do that, you can do spectral analysis on each pixel to see how much they differ from the surrounding terrain. Since we know pretty well what the typical composition of rocks on Mars is and what their spectral profile would be, we can look for wildly aberrant points, like polished metals that shouldn't be found on the surface. You can do this without filtering noise, but you get a lot of false positives.

Now, this transformation doesn't work very well with only 3-7 wavelengths of light like in most common image data, but it's just an example of things you can do to remove noise in satellite image data. There are other techniques that might work better with this data, but those are for people better educated in the fields of remote sensing than I to describe. (I'm just a code monkey who writes software for this kind of thing based on docs handed to me by the PhDs here. I just understand it well enough to implement it.)

Here's an image of the beagle

[AtariAmarok]

Please click here.

"We can find it if we know where it is!"

[elendel]

From the article:
"It would be extremely difficult to find a lander for which the location is uncertain"

So basically they are saying if they know exactly where the lander is, they can point to it on a picture of mars. The only real news is the picture has enough resolution you can see the dot that is the lander. Of course, if they are wrong, it might be that dot over there... or maybe that one... or...

That said, this is pretty damn cool. Hopefully the next mission will have even better resolution cameras, and we can get a better view.

Re:"We can find it if we know where it is!"

[Gogo Dodo]

That said, this is pretty damn cool. Hopefully the next mission will have even better resolution cameras, and we can get a better view.

The Mars Reconnaissance Orbiter launching in 2005, arriving in 2006 will have a higher resolution camera called HiRISE. 1-2 meters/pixel.

Re:"We can find it if we know where it is!"

[elendel]

Uh, this is the same resolution as the currenct MSG. From the article:

"Normally it can resolve features only down to about 3 feet (1 meter) per pixel, not good enough to discern a typical landing craft from its surroundings."

They got the additional resolution needed to see the landers by using a funkey trick, something about pitching the spacecraft at a faster speed... not very well explained in the article. But it gives a single-dimension resolution of ~20 inches/pixel.

The MSG actually only has a resolution of 1.5 meters/pixel, if the information here is correct. So The Mars Recon Orbiter camera will be a bit of an improvement, if they use the same resolution-enhancing trick. But it will only make the dot a little bigger... I doubt it will be able to confirm that the dot actually is a lander. Although, the HiRISE will have a 1-2 ft/pixel resolution at near-infrared. This could be be enough to almost make out something...

Re:"We can find it if we know where it is!"

[elendel]

Err, that's MGS, Mars Global Surveyor. We aren't sending any food additives to Mars, yet...

Re:"We can find it if we know where it is!"

[zero_offset]

Says here the MRO will have 20-30cm resolution...

Re:"We can find it if we know where it is!"

[jx100]

ooo.. Snake on Mars...

WMDs

[mraymer]

You can clearly see that the blob of pixels in question is, in fact, a Martian mobile weapons lab.

I think we are a few artist renditions away from a Martian Liberation.

Re:WMDs

[Bowling Moses]

the Martian Operation to Restore Equality, Opportunity, and Institute Liberty?

Re:WMDs

[iggymanz]

Marvin the Martian denied this saying "The rabbit lies. There is no Illudium Q-36 Explosive Space Modulator, there never was an Illudium Q-36 Explosive Space Modulator".

Not real - just a game!

[Anonymous Coward]

Pin the Beagle on the blurry satellite picture!

That's not random pixel noise ...

[fygment]

... that's my wife. Just kidding.

It's actually Saddam's weapons of mass destruction. Clever bastard! Just kidding.

I worked with imagery analysts in the pre-digital age. You wanted to see a field of camouflaged T-64 tanks in your picture? No problem. Change your mind and want instead to see the Soviet fleet exiting harbour at battle stations? Heck, it could be that too! They'd just change the captions. It was a standing joke. They worked hard and honestly I'm sure, but only they really knew the difference.

New trick?

[bobbozzo]

they used a new trick developed last year to squeeze more detail out of the camera

Why don't they just use the equipment all the police & spook shows/movies use to enhance crappy video so well that you can zoom into a 4-pixel person until you see the pores on their face??
:p

Better resolution image.

[Executive Override]

Better resolution image here.

When will we send higher-resolution cameras?

[johnjay]

Does anyone know if there are any plans in the works at NASA, ESA, or elsewhere to send higher-resolution orbitting cameras to Mars? I know that the current pictures are by far the most detailed we've ever had, but someone must be thinking of doing better.

They must have have a pretty good idea where the Beagle should be, a good enough camera might at least be able to verify that it's there. (Not that it would matter, but it'd be nice to know...)

Re:When will we send higher-resolution cameras?

[Gakster2001]

"In 2005, NASA plans to launch a powerful scientific orbiter, the Mars Reconnaissance Orbiter. This mission will focus on analyzing the surface at new scales .... For example, the Reconnaissance Orbiter will measure thousands of Martian landscapes at 20- to 30-centimeter (8- to 12-inch) resolution, good enough to observe rocks the size of beach balls." If I read that correctly, this will have tremendous optics. Read it here

Re:When will we send higher-resolution cameras?

[johnjay]

This is my third attempt at responding, slashdot keeps crashing.

Thanks for the info and the link.

In related space news...

[stendec]

According to the United Press International, President Bush will propose changes to the U.S. space program that includes a manned return to the moon in 2013 with an ultimate aim of a landing on Mars. Further plans involve retiring the space shuttle fleet once the ISS is completed.

Huh?

[DigiShaman]

I don't get it. What's so funny about that pic in relation to the parent post? Or is this more british dry humor that I'm oblivious to?

Beagle to enter final communication mode soon

[Inflatable Hippo]

After approximately 25 sols Beagle 2 will go into it's final communication mode SHM1 (Spell Help Mode 1). In this mode it
will shred it's own parachute and lay the pieces out across the martian surface to spell the world "HELP".

Calulations suggest that each letter will be approximately 4-6 pixels across, easily discernable from the Mars Express orbiter.

High winds may hamper this last ditch attempt for Beagle 2 to communicate.