Printable Version of Topic

Click here to view this topic in its original format

Unmanned _ Image Processing Techniques _ Registax and MER images

Posted by: Stu Aug 29 2010, 06:04 AM

After James mentioned "stacking" in one of his posts in the 'Distant Vistas' thread, I thought I'd have a go at using the astronomical image stacking freeware program "Registax" with MER images, just to see if it would work - hoping, of course, to bring out details on the Endeavour hills - and was quite pleasantly surprised with how it turned out. Others can do much better, I'm sure, but for a first go, hmm, quite happy.

Took half a dozen Pancam images of our old friend from Sol 2024 (Oct 2009), like this...

... and after stacking them together, and doing a bit more tweaking with contrast and brightness and cropping, got this...

As usual with my stuff, not claiming any scientific value, or new insights into the associated geology, just trying to make an attractive picture. smile.gif

Some more here: including one I'm really pleased with...

Posted by: Ian R Aug 29 2010, 06:48 AM

Well, you've succeeded with flying colours, Stu!

Posted by: elakdawalla Aug 29 2010, 02:40 PM

Stu, are you starting with the raw JPEG or with the archived data? You'll get better results with the latter, and it's available for that sol.

Posted by: Stu Aug 29 2010, 03:15 PM

Starting with raw, for now. Will have to build up to the better stuff, when I have more time; I'm very much an enthusiastic amateur with this stuff smile.gif

Posted by: elakdawalla Aug 29 2010, 05:31 PM

Using the raw images when the calibrated stuff is available is like being allowed to use anything, any paint, any canvas, any material in an artist's studio for free and saying, "no thanks, I'll stick with my Crayola™ 8-color watercolor set, thank you." The raw images are out there to pique your curiosity; you've been gaining experience with them for six years, and you clearly spend many hours a week manipulating them. You -- (and I'm not just talking to you, Stu, I'm talking to everyone who's reading) -- should conquer your fear that "it's too hard" because it's NOT too hard, there are so many tools available to help you get to the real data. Only showing images generated from the cruddy JPEGgy overstretched raw versions does a disservice to our ladies' amazing vision. Especially if you are going to go through processing steps like stacking -- for goodness' sake, don't handicap yourself with the JPEGs, use the real data!!

(This rant will no doubt turn into a blog entry this week.)

Posted by: James Sorenson Aug 30 2010, 02:23 AM

I certainly agree with you Emily. Lately I have been playing with the PDS data myself. The JPG's are a nice preview, but nothing compares with the original data from the PDS. Not just with MER, but with any mission available.

Posted by: Astro0 Aug 30 2010, 06:12 AM

I agree with the above, but it's worth noting that what Stu is doing is showing we lesser mortals another technique to get the most out of the stretched JPGs until the uncompressed stuff arrives on the PDS. Not everyone wants to wait 6 months until that data is released. A quick preview now using this stacked process is valuable and will help some of us who have less time and download limits to do something of value with the available data until the best stuff hits the PDS. I think that this is what UMSF is all about. We sit at our computers agog at the stuff produced some some of the better image mages we have here and wish we could do that stuff ourselves. Each and every little step and technique demonstrated on this Forum helps the rest of us get a little closer to the stars. smile.gif

Posted by: Stu Aug 30 2010, 08:00 AM

Well, I was quite pleased with those efforts, sat here at the back of the class with my crayons... sad.gif

I do know what you mean Emily, but I haven't used the pds stuff because, to be honest, I'm not a genius on a computer and I'm a lousy 'data trawler', so I stick with what I know I can do reasonably well, and am happy to let others with more technical savvy produce the masterpieces. That's not being lazy or scared, it's just knowing my current limitations. I really do do it just for fun. I don't feel any obligation to join in some image processing arms race. My flag is planted firmly atop 'Pretty Picture Hill' and I quite like the view from up here smile.gif

I hope no new members, or existing members, thinking about finally starting image processing, have been put off by this discussion. Yes, the raws are - well, raw, but you can still make great pictures out of them, and there's great a sense of achievement making a colour pic from a trio of raws. The images I've made are obviously not as good as they would have been if I'd used pds images, and I will probably try that eventually, but they've been very well received by the audiences of the Outreach talks I've given in universe knows how many school classrooms, community centres and village halls over the years. So, if you're sat there thinking about giving it a go yourself, then give it a go! And remember to have fun while you do it! smile.gif

Posted by: elakdawalla Aug 31 2010, 10:31 PM

But you don't have to be an "experienced data trawler," Stu, that's what I'm trying to say (and it may seem I'm picking on you, but I am talking to everyone who's gotten into space image processing because of the raw data but is scared to plunge into the real data). I just asked myself the question: "how can I get to PDS versions of the images in the fewest number of clicks, without any special software?" Please just humor me and follow my clicks:

Go to the
Click on the link at left to
Under the "Select Mission" window at top left, click "Mars Exploration Rover."
Under the "Image Type" thingy in the center, click "Regular."
Under "Planet Day Number," type in the sol number you're interested in (in this case 2024), same in both boxes.
If you want, you can select just Pancam or whatever, but you don't have to.
On the left, under "Product Search," click "Get Results."
There, now you can find all the images from sol 2024. Click on any one of them, and you'll get the option to download it in whatever format you like -- JPEG, PNG, TIFF, or GIF.

That's it. That's all you have to do. Four clicks and type in the sol number, and you're at the pictures you want. No special software. No programming. Nothing scary. The computer won't bite you. You do so much outreach and so much work with these photos -- I can't understand why you'd want to work with the crappy JPEG photos -- and show all the people you talk to those photos -- when just four clicks will get you to the REAL data, in nice, safe formats you're familiar with, like PNG. We work with the raw JPEGs because that's what we have to do to follow the mission in real time, but once the real data is available, with all the artifacts calibrated out, it's a terrible shame to keep working with the JPEGs.

And guess what? A similar series of clicks gets you straight to science quality photos from nearly every mission that carried a camera into space. Cassini, too, and Voyager, Galileo, Viking, etc. etc. Three clicks just got me to every single photo of Deimos from Viking Orbiter. It is all there at the Imaging Node.

JUST FOUR CLICKS. EVERYBODY DO IT. I know it's intimidating, so do it when nobody else is watching. But there's no reason you can't look at the real data. Don't sell yourselves short, you're all scientists, so look at the science data!

Posted by: Stu Aug 31 2010, 10:44 PM

I'll be totally honest, I had no idea it was actually that easy. I have tried accessing pds data before but got lost on the page I tried. Looks like it's time to give it another go. smile.gif Thanks for the step-by-step guide, Emily, that will be extremely useful for a lot of people here.

Posted by: elakdawalla Aug 31 2010, 10:53 PM

For some missions it's not *quite* that easy -- for instance, they archive HRSC and MOC data, but not in easy formats like PNG or TIFF, for those you have to be able to handle IMG. But there is a shocking amount of data available in just those few clicks.

Posted by: Stu Aug 31 2010, 11:27 PM

...and, my first PDS-derived image...

Looks like I'm going to need a new external hard drive... blink.gif

Thanks Emily, it's going to be fun working with that data.

Posted by: elakdawalla Sep 1 2010, 01:00 AM

Dragged you kicking and screaming, but I got you there biggrin.gif

Posted by: Astro0 Sep 1 2010, 04:15 AM

OMG - Stu will be impossible to stop now.
I can hear him drooling all over the PDS from here ( eww! tongue.gif )

Posted by: Stu Sep 1 2010, 06:08 AM

QUOTE (elakdawalla @ Sep 1 2010, 02:00 AM) *
Dragged you kicking and screaming, but I got you there biggrin.gif


No, seriously, I was under the genuine impression that it was a LOT harder to get at and use that data than that. "PDS data" has always been spoken about in hushed, almost hallowed tones, and I'd been to that very page before and got absolutely nowhere, so I had comvinced myself there must be a lot more to it. It's never been explained so simply and practically before, so thanks again. smile.gif

Posted by: djellison Sep 1 2010, 06:19 AM

If you want to go slightly more hardcore - you can download IMG's and process them VERY easily with Bjorns little program. I say it many many times, but it's true - I'm not a code monkey. I can't do HTML or PHP or any of those acronyms of hell, but even I... a temperamental artist (and I'm not even a very good artists ) - can manage that smile.gif

It's VERY satisfying to go from unintelligible files, to a well stitched colour mosaic.

Posted by: Stu Sep 1 2010, 06:35 AM

Give me a chance; I've just figured out how to operate this Stargate... laugh.gif

Posted by: climber Sep 1 2010, 08:09 AM

I love this discussion! It's exactly what UMSF is about.
It reminds us we have to fight to get results and there is always somebody here to help... and challenge.
Thank you

Posted by: ugordan Sep 1 2010, 08:26 AM

QUOTE (Stu @ Sep 1 2010, 08:35 AM) *
Give me a chance; I've just figured out how to operate this Stargate... laugh.gif

My God, it's full of... images?

Posted by: Stu Sep 1 2010, 02:02 PM

To support Emily's point, here's a quick but eye-opening comparison between images made with, on the left, the "cruddy" ( © Emily Lakdawalla ) wink.gif raw images and, on the right, the equivalent PDS images...

Looking forward to trying this with some Endeavour horizon shots...

Posted by: hendric Sep 1 2010, 08:16 PM

I was about to say "Now THAT looks like an iron meteorite!" The images are significantly better. Great job!

Posted by: fredk Sep 3 2010, 06:13 PM

Just a niggly comment: I would call the archived data "raw", not the jpegs available in realtime on the jpl site or exploratorium. In digital camera lingo, the raws are uncompressed, and therefore best quality, while the jpegs are compressed and so have lost information.

Posted by: elakdawalla Sep 3 2010, 09:02 PM

I agree that that's how the terms should be used, but since the Cassini and MER websites use "raw" to describe their JPEGs, that's the way the terms are going to be used, I'm afraid.

Posted by: Stu Sep 11 2010, 09:17 PM

Some more PDS-derived images here...

Posted by: MouseOnMars Jan 13 2011, 01:47 AM

Is the stacking technique recommended for vista/horizon images ...

If its possible to get a clearer more dramatic view across the plains to Endeavour then I'll give this technique a shot.

Powered by Invision Power Board (
© Invision Power Services (