h a l f b a k e r y
Sugar and spice and unfettered insensibility.
idea:
add, search, annotate, link, view, overview, recent, by name, best, random
meta:
news, help, about, links, report a problem
account:
Browse anonymously,
or get an account
and write.
or Create a new account.
|
|
|
I am positive people have thought of this one before, but I don't think it exists quite in the form I'm thinking of.
The idea is to have software that uses pattern recognition on a night sky to determine geographic position. This is running on a laptop coupled to a high resolution digital camera with
an extreme wide angle or fish eye lens. The camera is facing up.
This will be independent of the GPS.
The software will recognize constellations and other celestial bodies and use the information to get a fix.
I think systems like this very likely exist in military applications. I found one reference to a guidance system from the 1960's that tracks stars, but I think this isn't quite the same: I think that system follows just a single or a few stars, but doesn't actually get a fix out of nowhere.
Like this?
http://www.calculat...ce.com/st-sppc.html
[21 Quest, Oct 30 2006]
Stellarium
http://www.stellarium.org/
Really sweet free planetarium software. [jmvw, Oct 30 2006]
John Harrison
http://en.wikipedia.../wiki/John_Harrison
This is the chap CustardGuts is talking about - Harrison did win funding from the Longitude board, but died (aged 83) before ever receiving the £20,000 prize. [zen_tom, Oct 31 2006]
[link]
|
| |
It sounds possible if the sky is largely clear, but what about clouds or light pollution? |
|
| |
They would impair its performance. |
|
| |
"My PC's got no legs."
"How does it navigate?"
"Celestially." |
|
| |
[zen_tom], stop messing with my head! [21 Quest], nope. That doesn't do the pattern recognition. |
|
| |
One day I will make this. If I want to. |
|
| |
my understanding is that if you know the time, you can work out where you are (east/west), as long as you can see the horizon. OR, if you know where you are, and can see the horizon, you can determine the time. it's one or the other. |
|
| |
If you can see the horizon, you can know where you are north/south. |
|
| |
The stars are so far away that there are no reference points (except when using rediculously accurate telescopes), and it is the horizon that you must measure against. |
|
| |
I'm not sure if it's still available, but at one point there was a prize available from the british government; basically if you could determine lattitude (east west position) to within a certain level of accuracy, you'd win a humungous prize. This was because of difficulty navigating ships at sea. All sorts of people tried, to varying degrees of success. eventually, I think the guy that invented the spring wound clock, or some accurate variation on it (a pendulum clock doesn't work as well on a moving, rocking ship) won the prize, but I'm not sure. For some reason I seem to recall that because it relied on a clock being accurate, they didn't qualify. |
|
| |
It's really a rather challenging paradox, if you think about it. there are no reference points. If you don't know the time, date, and a heap of astrological information, you simply cannot determine lattitude without a GPS. I read a book about it once. |
|
| |
Perhaps the prize is still available? |
|
| |
That's a terrific thought. I'm not about to win that prize. We'll include a high quality clock and an artificial horizon: a glass ball filled with a drop of mercury and lots of little contacts inside. I removed the remark about determining both time and location. |
|
| |
The prize is for determining longitude (position east, or west) Lattitude can be determined with pretty decent accuracy based solely on the stars present. Clocks allow you to determine Longitude because you know what time it is where you left (from the clock) and you can determine the time where you are. |
|
| |
It is also possible to determine longitude (or time) based on the distance between the moon and various stars. The process of performing a "lunar equation" to determine this was an important detail discussed in the Nautical Almanacs well into the late 1800s. Be sure to include that in this system, in case this enormously complicated computer system needs to account for a broken clock. |
|
| |
That's really good, ye_river_xiv. The moon. |
|
| |
The planets may also be close enough to give us accurate information, if they're near in orbit? |
|
| |
I for one think this is a top idea, especially the Optical Celestial (Body) Recognition. |
|
| |
sorry, got the terms longitude and lattitude mixed up. I meant north/south is easy, east/west needs a clock. |
|
| |
Artificial horizon. Awesome idea, well done j. |
|
| |
still needs a clock, though. |
|
| |
the moon idea is ok, but once again i imagine it would need a clock, too. |
|
| |
hhhmmmmm. many people, much smarter than I have pondered this one, methinks. |
|
| |