Found In Space (On Flickr) 48
Jamie stumbled upon a writeup for all you astronomy and photography buffs out there (Perhaps my Dad or Uncle Jim are reading ;). From the writeup "The 'blind astrometry server' is a program which monitors the Astrometry group on Flickr, looking for new photos of the night sky. It then analyzes each photo, and from the unique star positions shown it figures out what part of the sky was photographed and what interesting planets, galaxies or nebulae are contained within. Not only does the photographer get a high-quality description of what's in their photo, but the main Astrometry.net project gets a new image to add to its storehouse of knowledge." Check out the Astrometry.net site for many cool pictures.
Astrometry.net and Comet Lulin (Score:5, Interesting)
I use them all the time; just shot comet Lulin [flickr.com], they did a great job of exact location:
Re: (Score:3, Interesting)
I remember the first time that, just using a consumer-grade, non-DSLR camera (a Sony DSC-H2), I was able to image the Galilean moons. Blew me away. At first I thought I must have just captured bright background stars that just happened to be near Jupiter in the same plane, but I went back home and set my time and location in Celestia, it was a perfect match. I took more shots later and was able to see them changing positions around Jupiter appropriately.
It's amazing what you can do without any special eq
Re: (Score:3, Interesting)
As far as I'm concerned with the current crop of DSLRs, we've entered a pretty darned fabulous era of shooting at night in general and the night sky in particular. It's getting to the point where they're trying to squish just the faintest remaining noise in the row/column amps and leakage from the sensels. It's already in the blinking amazing zone, but Canon seems to think they've got another full stop of noise reduction up their sleeve; all the online tech folk are doubtful, but me, I'll wait on the engin
Find any killer asteroids yet? (Score:5, Insightful)
It's only a matter of time before The Goog aggregates all pictures of the sky in realtime to find killer asteroids & make weather forecasts.
Re:Find any killer asteroids yet? (Score:5, Funny)
It will probably find evidence of vast machine intelligences like itself and decide not to tell us.
Re: (Score:1)
-- Ads provided by Google --
Have you ever felt like your body was a mere vessel to help you in your journey? Do you like Kool-Aid? Maybe Heaven's Gate is right for you. Click here to find out more
Photosynth for the sky? (Score:2)
Is this just Photosynth for the stars, finding common points and linking the images?
Re:Photosynth for the sky? (Score:4, Interesting)
It's similar to Photosynth in that it finds the stars in the sky, but it only provides location information. It does not provide links (you can't navigate from photo to photo the way you can in photosynth.) Not to say they couldn't add that functionality later.
I think of this more like a real-life version of the "Astrogator" role on the space ships from old sci-fi stories, where they arrive at some spot deep in space, the astrogator looks around at the stars and determines exactly where they are.
Lost in Space sequel? (Score:2)
This is also the way I understood TFA.
Very cool, IMHO.
Re: (Score:2)
Very much like the astrogator, with one minor draw back. It only works from one position.
Re: (Score:1)
I guess, in theory, if you work from a single reference image that you know the exact date and time, and proportions (for measuring shadows) you could create a 3D world that was made from a mesh of user's photos. As long as you have the date and time (within a minute or two) and your image was connected to another image that was ultimately connected to the reference image (for scale) you should be able to calculate your lat and longitude based on shadows, knowing where the sun is in relation to the earth a
Awesome (Score:2, Interesting)
Re: (Score:3, Insightful)
May I also posit that it not only keeps the spread of information flowing, but increases both the available information to each of us and the aggregate total information available to mankind.
Small changes like this change how we see life in the long term. Seeing the sky as if it were so many maps of a foreign country brings with it familiarity that raises awareness and new thinking.
This, and projects like it, are awesome. Spell that with capital letters and boatloads of enthusiasm.
Re: (Score:2, Funny)
Re: (Score:2)
ROFL... ok, spell awesome with capital letters and boatloads of enthusiasm
Re: (Score:3, Funny)
I wonder if they'll be able to figure out where I live though...
Given that they're matching these pictures against those taken over the last 60 years or so, and considering the accuracy of the typical consumer camera, I think they'd be able to pinpoint you to an orbit somewhere around the star Sol.
Very cool, but np-complete? (Score:2)
Re: (Score:2)
The "interesting points" are obvious: dots on the picture. You take the brightest ones, and then search for the pattern generated by those dots in your database.
Astronomy actually seems like one of the easiest applications of this kind of thing.
Re: (Score:3, Interesting)
There's a PDF link buried in the interview [nyu.edu] which had me giggling with geekish delight. The phrase
makes me happy.
Also, I'll give credit to waxy.org [waxy.org] which is where I read this first.
Re:Very cool, but np-complete? (Score:4, Interesting)
If it is done in polynomial time, then it would by definition be in P which means it would NOT be NP complete unless P=NP.
I think you are trying to ask how it could be done in less than exponential time (which is how fast the fastest algorithms to solve NP complete problems run).
I personally would expect an algorithm to perform the match would run in O(U*US+K*KS+UI*KI*U) where U=the number of unknown (ie flickr pictures), K=the number of known pictures, US=the average size of an unknown picture, K=the size of the average known picture, UI=the number of interesting things in an unknown picture and KI=the number of interesting things in a known picture. However, it would often run faster because it should be fairly easy to disqualify two pictures as containing any of the same interesting features.
Re: (Score:2)
If it is done in polynomial time, then it would by definition be in P which means it would NOT be NP complete unless P=NP.
No, the N means it can only be calculated on a nondeterministic (Turing) machine in polynomial time. Common mistake.
Re: (Score:2)
Re: (Score:2)
Question is, is GP also a superset of NP ...
Re:Very cool, but np-complete? (Score:5, Informative)
You can think of our search as randomly trying millions of possibilities and hoping for the best. The reason it's so fast is that we order the random attempts very cleverly and tend to find the answer fast if it is indeed solvable. The algorithm usually terminates because it finds a match or times out; rarely does it exhaust the search space in time. The actual complexity of our system is roughly O(N choose 4) where N is the number of stars in the image. Interestingly, this is polynomial, roughly O(N^4), though probably closer to O(N^5) once verification is added.
In summary: the astrometry problem is not NP-hard when approached like we do.
Disclaimer: I am one of the astrometry.net contributors.
Re: (Score:2)
Drom, very interesting post. it seems to me that you guys are using some kind of heuristic to solve the problem. Which I find amazing. Can you delve a bit on it, or would it be too problematic? I got really curious about the heuristic you guys are using.
Re: (Score:3, Informative)
I'm guessing its basically just a star-tracker algorithm. I've taken a class on it, but its probably the thing I understand and remember least since I've been in grad school. Basically, they pick out the bright stars, and measure their locations to sub-pixel precision through centroiding, so you're working on a set of coordinates, not an actual image.
Then an algorithm cycles through a star catalog to compare the relative positions and identify the stars. In this case, since the angular size of a pixel is
Re: (Score:2)
It probably can be shown to be equivalent to the Clique problem.
Maybe some approximating could do the task in P too, but that doesn't say anything about the constants in your Landau-estimation. Your NP algorithm may be way faster than your P algorithm ...
The illuminated. (Score:2)
"Not only does the photographer get a high-quality description of what's in their photo, but the main Astrometry.net project gets a new image to add to its storehouse of knowledge."
Dear Astronomy Buff. You have an excellent picture of a cruising UFO.
sounds like a good time (Score:4, Funny)
Re:sounds like a good time (Score:5, Funny)
to confuse the shit out of someone with photo of a backlit piece of black card with random pinpricks in it.
Sounds like something a random prick would do.
Re: (Score:2)
To be fair it raises a point though- if the system uses this data automatically then can it tell the difference to say an image from a screenshot of a game like Eve online? I'm not sure what the data they're receiving is actually used for.
What if a user has a rendered image on their system that they downloaded as a background or something and uploads it believing it's a real image?
Re: (Score:2)
They already have an independent catalog of star data (the USNO-B) from which they built their reference search data. It's used to locate the incoming pictures, and they won't extend the catalog from user-submitted pictures.
However, since they're talking about using the data to locate faint moving objects, it would be theoretically possible to submit a series of doctored photos that show a fake "object" moving.
Even this would be pretty much useless because if you faked anything of interest, such as an
Correction (Score:1, Insightful)
Check out the Astrometry.net site for 4 cool pictures.
Am I missing something on the gallery page?
conspiracy (Score:1)
Moar DATA (Score:2)
This will get even better once Pan-STARRS [wikipedia.org] and LSST [wikipedia.org] get up and running
Whoo would have known? (Score:5, Insightful)
When these kinds of services (meaning flicker et al) open up their api to public use there are always some who cant figure out why.
Well folks, this is the reason why.
I betting when the executives at flicker sat down to decide if they were going top open up the api, they had no idea that someone would use it to create a map of the sky.
Re: (Score:1, Funny)
Major Discovery (Score:2, Funny)
Article summary raises interesting questions. (Score:1)
Like how do you properly post an emoticon in () text? http://xkcd.com/541/ [xkcd.com]
Re: (Score:3, Interesting)
Use a different font. Slashdot makes available the <tt> tag which works nicely - emoticons often look better in monospace text anyway. I suppose that's probably subjective though; I grew up with monospace text, if you didn't you probably feel the other way. I think we can all agree that people who use the tag and the enter key to format their text as if they were old school are extremely fucking annoying, though.
WOW! (Score:1)