Curiosity's Cap 1B phase was completed yesterday with the successful first drive, so now we're going into "Intermission." Dan Limonadi's guest post on the Society blog has great explanations of what all these phases mean. Keep discussion of sol 9-16 imaging in the relevant thread -- not all of those full-frame Mastcam images are down yet.

First sol 17 Navcam thumbnails are up on the raws page -- more views of Mt Sharp.

Could someone please de-Bayer this:
http://mars.jpl.nasa.gov/msl-raw-images/ms...2000C0_DXXX.jpg
And pass the swear jar!

With pleasure



And with a bit more levels :



The details are… I have no words for now

Here's my attempt at debayering. I cooked it pretty hard to see how much detail I could get out. The patches of green and purple are not real, those come from JPEG artifacts in the source images.

Click to view attachment

Looks like a vintage postcard Really stunning.

Look at the little sand drifts in the near horizon on this one! http://mars.jpl.nasa.gov/msl-raw-images/ms...3000C0_DXXX.jpg

Yes Emily. And I'm wondering if Curiosity's team is planning to open an Instagram account to the rover .

This view, in color :

These are amazing pictures... Fred B - absolutely stunning! It's going to be incredible driving around that topography.

But first - Glenelg. I hope we'll see another drive soon. I will be posting route maps from time to time, but not after every drive. In particular I'm not able to do so from Aug. 31st to Sept. 16th, but look for something after that.

Phil

Wow, I think we can see there is nothing wrong with the focus of Mastcam100. Those lossless frames are really something, too bad they are not released losslessly.

Lovely description, brilliant results!

A drive was planned for 23rd but postponed (Paolo via Twitter) - maybe tosol?

Phil

Hi to everybody,

This is my first try at converting the color information on the right MastCam images. Altough is not as good as I would like (don't know if it's due to JPEG compression or because I don't know how to do it better) it clearly shows the color of the soil and the rocks. The ventifacts are pretty nice

Click to view attachment

Comments and advice are welcome!

Greetings!

That looks like the best yet!

So what is the deal with the bayered images? Some mastcam images are de-bayered on Mars (or back home) and some are not? The algorithm for Mars scenes should be easy--just assume a shade of red & you're good.

Sure would be nice if Adobe Lightroom had a "MASTCAM RAW" plugin. Of course, jpegging them beforehand is sub-optimal.

Sol 21, but things might change (again).

Paolo

The ones that are returned in original, Bayered form are lossless which is A Good Thing.

Every new frame is a surprise for me. With the new raw frames from right MastCam downloaded some minutes before... we have a vertical panorama!

Click to view attachment

This is the little version. If you want to see the full version of the panorama, go to: http://www.ungeologoenapuros.es/wp-content...XX-Panorama.jpg It's 8 MB.

Comments and suggestions are welcome!

An other unbayered image.



There is not only Photoshop in image processing life, I'm actually using ImageJ with the "debayer" plugin, available here (scroll down). Then, in ImageJ set the image in 16bit mode. Then, go to "plugin>debayer" (you should have install it in the correct folder in order to have it in your plugin list). The best method is "R-G-R-G", set Demosaicing Algorithm to "Adaptive Smooth Hue" (I think it gives the best results). Checkbox "Display color image". Et voilà ! Save or do whatever you want with your debayered image .

Great images, especially the vertical panorama!

I'm puzzled why we are still seeing hue quantization in these de-Bayered images, as we did in the color JPEGs. Why would JPEG compression of the Bayer pattern produce that effect, rather than just randomly perturbing the colors?

John

Since the checkerboard bright/dark appearance of the Bayer pattern in grayscale depends on the original color of the scene, various blocks can end up having different DCT frequency domain component breakdowns, but not all frequency components are quantized equally. This, I think affects the amount of pixel crosstalk. If one overlays a Bayered image over a color one, one can see the places that experience color shfts have a distinctly different checkerboard pattern feel to them. Not completely the same mechanism as banding in highly compressed color images, but I guess you can say the root cause is the same.

I like this M-100 shot:
Click to view attachment

It's even possible to achieve a good, but not perfect, combination of MastCam left and right to make some anaglyphs, but the resolution is not as good as it could be due to the difference of angle of vision.

Click to view attachment
Click to view attachment
Click to view attachment

That's beautiful.
I really think the right 100mm camera won't stop providing us stunning shots.

That vertical panorama is terrific.

I have no sense of how large the pebbles are in the foreground -- I think they may be quite tiny. We need the Mystery Man -- and maybe the Mystery Man needs a Mystery Mouse companion to show us the size of those pebbles!

There are some cleat-marks in the soil midway down. What's the cleat-to-cleat distance on a MSL wheel?

Also interesting to note one rock sinkage after been roved over

Yes, great work Parai'

My own version (full res available by clicking the image) :



And there is also a M34 pan, incomplete for now, I think.



The M100 vertical pan is pretty large in vertical. Imagine now a full 360° acqured with this camera. No, my computer will die way a lot before finishing stitching only a half of the pano. A quick math :

The last M34 pan was 29 006 px wide for 3635 height. This column from M100 is 14 799 px, so about 14 800 px. A little cross product :

29 006 => x
3 635 => 14 800

y => 29 006 × 14 800 / 3 635
y => 118 098

So, this will result in a panoramic of about 118 000 pixels wide !! This is HUGE ! Is there computer which can just open such a huge picture ? I pass my hand…

Thanks for that great deBayering tip Ant103! Here's a very simple stitch in the meantime.

Click to view attachment

Well, well, well... turns out Registax has a DeBayer function...

Click to view attachment

Now all I have to do is work out how to get rid of the terraformed hills, and the handfuls of emeralds scattered over the ground...

In reply to Emily's question judging by this NASA JPL image the Zig-Zag cleats look about 5 cm apart.
Roy

Easy - just host it at the Gigapan website

The grousers are 15 deg apart, therefore the distance is about 6.5cm.

Paolo

Here is my attempt at an anaglyph using ANT103's work and just playing around with StereoPhotoMaker. I just warped the hell out of those graphics. I see 3d but my eyes tell me they are being twisted in there sockets!

Click to view attachment




Note: It is possible that I used a medium resolution version of ANT 103's MastCam100's Graphic! The resolution could be much worse than it should be. I just clicked on the images and copied them to my clipboard. Then pasted them into IrfanView.

I would expect that most panoramas will be just the 34 camera, and the 100 camera will mostly be used for limited areas. But regardless of how they are used, what amazing cameras we have on this machine! Now let's see some micro-images...

Phil

Another deBayered stitch.

Click to view attachment

Matt Heverly‏ tweets that the next drive is tomorrow after a quick detour to one of the scour marks to make an observation (Chemcam, presumably) before driving east.

Phil

We will be placing DAN on top of the northern scour mark.

Paolo

My best shot at a greyscale of the mastcam 100 view of the mound. I took the two slightly-different pointing frames, applied my Fourier filter to remove the jpeg Bayer artifacts as well as I could, de-Beyered, then averaged the two frames to remove even more residual jpeg artifacts, finally a contrast stretch. In the end, I think most of these details are real!
Click to view attachment
Just think: that's where we're going!

Another try at a mastcam anaglyph. Here's a smaller version of the ML image with the MR area shown by the rectangle:

Click to view attachment

The anaglyph:

Click to view attachment

Fredk : I've took your fantastic image and applied some wavelett sharpen. And then add color from one of my debayerized pics. Hope it's okay
Click to view attachment

Uh...oh, my!

Wow.

The layering blows my mind.

This should probably be in the Making anaglyph images, Methods and software thread but my question is did you have any further adjustments in size after cropping out the image from the larger MR area? I did somewhat the same thing*, but the objects in the ML image still appeared much smaller. It looked really weird until I started playing with something called barrel distortion in StereoPhotoMaker. I don't claim to know what I am doing but if I try enough things sometimes something works. If I don't need to correct for barrel distortion that would certainly make things simpler.

* I didn't even use the full ML graphic.

I finally got this. ( click below ) Comparing mine to yours - yours certainly looks cleaner. However; mine may have more 3d effect in the fore & middle ground? Yes? No? Maybe?
click link:
http://www.unmannedspaceflight.com/index.p...st&id=27692

http://www.unmannedspaceflight.com/index.p...st&p=189878


Jack

Fred -- if someone showed me your b&w image in almost any other context and asked me where the picture was taken, I'd probably say somewhere in the canyon regions of the American Southwest.

I'm with Nick -- absolutely blown away.

This is gonna be *fun*, guys....

-the other Doug

Enhanced Fred's greyscale a bit

One more try to really bring out some details... I can't wait to get closer to those hills!

Will the Mastcams still need to be formatted in this way once they reach the PDS, or is it just because of the fact that these are JPEGs?

It seems that during Sol 19 a relatively loose rock or soil has (maybe the wheel tracks?) been shot 5 times with the laser leaving a little trail of holes:

Click to view attachment

I hate to be first to say this but...

pew pew pew pew pew !!!!

.

I borrowed Fred's excellent rendition to compare with HiRISE anaglyph of the prime science region around the inverted riverbed. It's a very narrow angle of view. If I marked some features wrong, please let me know.

Click to view attachment

The canyon between "A" and "D" will be epic, if they decide to rove up it. In Matthew Golombek's "Update on Landing Site Characterization" presentation, this canyon's traversability is marked in 4 places as "hasn't been investigated," and marked in 2 places as "small width path, less than 10 meters." Have they determined the traversibility since then, or will it be determined by the rover once it's in the vicinity? I would imagine the long lens in the Mast Cam could get some good shots from the base of the canyon and spot the rough patches. By the way, here's the page from the presentation with the possible routes:

Click to view attachment

I love the outcrop near to point B in the walfy picture. It looks like a plane of ripples

Also, as Gladstoner says, it seems to be two different distributions of rocks over the soil, and the little ones seem to be more rounded (the bigger ones are now being actively being eroded by the wind as we can see in it shape).

Just did it by eye in Photoshop--crop the ML, scale it up so that it matches the size of the similar area of the MR, apply level changes to RGB as appropriate, then fine tune the registration by eye using the Move tool and cursor keys until it looks about right. Lots of variables to mess up & it always seems like there are problem areas & as you say it gets a little brain-bending. But haven't tried correcting for distortions--maybe that's key. I suppose the pros use the full CAVHORE camera models to really do it right. Here's the best I can manage with one of the rover track pairs, without any geometric changes:

Click to view attachment

Looks like you might have stitched a couple of frames together? That's another complication to get right.

I don't find the two debayered frames. I would like to try my anaglyph version. Could you post the two frames? thank you