I’m not going to post an updated analysis image this week, even though the goal of >50,000 craters was reached. Instead, since we’re still in beta, I thought I’d spend a few paragraphs telling you what we’ve been working on behind the scenes here at Moon Mappers and CosmoQuest.
First, for CosmoQuest in general, Pamela Gay has been meeting with the advisory team as well as working on the Hangouts interface (not ready to beta this yet, sorry!).
The programming team, lead by Cory and a close second Joe, has been hard at work behind and in front of the scenes, too (you may have seen Joe on the forums under “Ansr” which I think is short for “answer”). There are two changes in particular that you may have noticed to the interface that are the result of these two gentlemen.
The first is that under the big blue “Done Working” button in your toolbar, Joe has inserted an “Image Link” hyperlink. If you click it, a new window will pop up that is a direct link to the image that you were working on. The main benefit of this in the short-term is that if you want to post it to the forum to get any feedback, you can do so without having to take screenshots and upload them.
The second – and probably most prominent – change is that every 15 or so images, you will be pitted against a crater expert (me) to give you a general idea of your accuracy — basic feedback. We’re still tweaking this and figuring out exactly how the comparison algorithm is going to work, but the basic idea is now there thanks to Cory on the backend and Joe on the front.
And Joe would like to know if anyone has been having issues selecting crater markings that are on top of each other. Please let us know on the forum or in the comments here.
In terms of content, the other co-science lead, Irene (“IreneAnt” on the forums) has been working extensively on writing up our Glossary of Surface Features section of the wiki.
Meanwhile, I’ve been working a lot on getting the analysis codes written. There’s a lot to taking everyones’ individual markings on a given 450 by 450 pixel image, reconstructing the Lunar Reconnaissance Orbiter camera image from each little image slice, and then clustering all of your individual markings into a reduced set of craters. But it’s not just that analysis I’ve been working on – it’s also figuring out how to figure out accuracy of all the different users, analyze the Man vs. Machine results, and working with all of the now >1000 features that you have marked across the different images.
So we’re still in beta, but still plugging along, making adjustments to hopefully make things a bit easer, getting code in place for future science, and other things.
And speaking of future science, I had a teleconference yesterday morning with the science team for the upcoming CosmoQuest’s Asteroid Mappers: Vesta Edition. So stay tuned!