View Full Version : Geomorphological Features

2014-Jan-26, 01:33 AM
Why are we not able to mark features such as landslide infills, rills, drag marks, or anything geomorphological. Given proper index of these features, the general public may help ID these features which may also help tell the story of the life of Vesta and soon Ceres.

It would also be advantageous to be able to see the higher resolution images. I really doubt the images captured are only 600 dpi x 600 dpi. This will help to zoom in and see grater detail such as bolders or any other such important features to note. Note only this, it would be nice to see a full 3D rotational image that we would be able to see and zoom into. Kind of like Google Earth's add on for the Moon and Mars instead of showing a small image each time to interpret. Sometimes seeing the whole picture helps to identify the history and smaller details better.



2014-Jan-28, 11:34 PM
Hey, bldrnnr11. I've passed your questions along to one of the project's principle investigators since they are better equipped to answer these. As I understand it, the features to be marked are the ones being targeted by the primary science goals of the mission team.

Will let you know what I hear back!

2014-Jan-31, 02:59 AM
Hello NoisyAstronomer and Bldrnnr11,

I agree with Bldrnnr11, these features (similar to Google Moon) would be great additions to the project; but I think that it would be hard to code it all, and out of the scope of what is needed for this project. At a minimum, even just seeing a zoomed out view of the image we are working on would be a great step, so we could get a sense of perspective.

Bldrnnr11, what I do to get a better look at the images is I download the image (right click on it and "save as", etc.), then open the image with an image processor (I use FastStone Image Viewer, which is free). Then, you can enlarge to full screen, sharpen the image, and work with the contrast, brightness, etc. to bring out more detail. Then I go back to Vesta Mapper, and looking at the same image I can see better what I am looking for (mostly craters and boulders, chains, etc).

In addition to seeing more clearly, this brings out the beauty of the surface of Vesta, in a way that is only hinted at in the 600x600 size images. I have even displayed a few on my TV screen, that were particularly stunning, to share with my wife.

For the interesting geomorphological features you mentioned, you can mark them as "unknown", which may draw attention to them.

Hope this helps!

Best Regards,


2014-Feb-07, 01:57 AM
At least for the Moon and Mercury, I can give you a bit more of our justification for the perceived limitations:

More Feature Types: While there are clearly people who know their geology and can identify dozens of different types of features on these bodies, there are many, many more who can't. Providing a very large selection list of different choices for what a feature is would overwhelm many people. So, for simplicity, we ave you the most common kinds of features - or features we were most interested in - to identify and everything else should go into that "Unknown" or "Other" category.

Zooming and Context: In a study we've been working on now for over a year (slow reviewers ...), the most common image manipulation that experts used that is not available through CQ is zooming. But, surprisingly, this inability to zoom did NOT alter results - at least in craters found and measured - relative to experts. Having the ability to zoom in and around would also take away from the simplicity of the interface: Do an image, done, move onto next. It lets you do things in bite-sized chunks and provides us with a simple metric of progress through our image sets.

Higher Resolution: I don't know the details of Vesta images, but at least for the Moon and Mercury, the vast number of images you see are full-resolution. The full image is chopped into 450x450 px sub-images, and then that's what you see. They are also reduced in size by 3x, chopped into 450x450 px again, and you will also see those. And then again. But, to give you an idea, for one of our Moon images, there are 186 full-res images, 21 reduced by 3x, and 3 reduced by 9x. So out of 210 images, 89% are full-res.

The guiding principle when developing these applications was simplicity. What is the minimal amount of "stuff" you need to identify the features that we want to study? What is the easiest way to create those tools for the easiest interaction? I kinda liken it to Apple's product line: For 95% of people, the iMac works fine and does everything you need. For 5%, they need the Mac Pro because there are just some things that you want to do that you can't do with the simpler iMac -- they're the "power" users. In my work, I'm a power user, and I fully recognize that the CosmoQuest interface may seem limited to other power users, but for the data we need to gather, the iMac works fine.

Hope that helps clarify some things!

2014-Feb-07, 05:47 PM
Wow, flya. I'd love to see it on a big screen TV. Maybe I can try that at home!

2014-Feb-07, 05:52 PM
Thanks, astrostu! Asteroid Mappers has the same design principles as Moon and Mercury, so a lot of that definitely applies.

I heard a bit from one of the Vesta PIs, and I think we went with the highest res Dawn data that was publicly available. And although we don't want to overwhelm the users with more detailed features, you can use the "unknown" or "odd" to mark something you think is important, as flya mentioned. There definitely is a balance that the designers had to strike, keeping in mind the wide range of users we wanted to attract, including school children through their teachers' use of the lesson plans we create along side them.

Thanks for the feedback!

2014-Jun-09, 11:27 PM
I have never jumped onto this project, because i cant scroll with the laptop mousepad.

I still don't understand why we shouldn't be marking up as many features as we can positively identify.

I know the PI is using this info for a focused project, but having the whole asteroid tagged with as many geo tags as possible opens up the data for more streamlined future studies.

I got frustrated with the Disk Detective for the same reason. There were 80-90% of the pics were seyfert I or II galaxies, and we had no way to mark em, except html tags.
What kind of statistical changes would this make for general cosmology ? Who knows, but felt like was just wasting time going to star catalogs, and then not making useful datasets out of them.

2014-Jun-16, 06:48 PM
morganism -- keep in mind that not every feature that you think is interesting is necessarily morphologically interesting/unique to a geologist. There's also the problem that if you tag a gagillion things, it's no longer useful. You want to call attention to the most interesting things, especially those that were asked for. We look at these images, too. It's when you call out one or two features per image that we can then make sure we pay attention to them.

2014-Nov-13, 10:59 PM
What we need is a "feature" marking to show tunnels. I have found at least one and have had to mark them as "odd-shaped features"

2014-Dec-11, 09:30 AM
Ajdunn - Any chance of seeing that tunnel image? Sounds interesting.