Classifying galaxies - citizen science

Hill · 08-10-2007

Hubble Deep Field Image. Every object here, aside from the two obvious intervening stars, is a galaxy.

I came across this neat site the other day called Galaxy Zoo Project | Galaxy Zoo

As they explain it :

Quote:

Astronomers have spent many decades trying to measure basic galaxy properties such as age, mass or dustiness that may give us some clues as to how they formed and evolved and what precisely the connection between spiral and elliptical galaxies is. However, most studies of galaxies so far have only looked at a few dozen or hundred galaxies in the nearby universe and many aspects of galaxy formation and evolution are still a mystery.

However, with the advent of the 21st century the age of large-scale astronomical surveys has arrived! The Sloan Digital Sky Survey (SDSS) is taking images of a large portion of the sky and will eventually find a million galaxies. With such a large number of galaxies, astronomers can finally begin to understand how they form and evolve by comparing various populations to each other with large enough numbers to draw real conclusions about their origin.

But out of these million galaxies, how do we know which are spirals and which are ellipticals? The answer is simple: look at them! Indeed, until now galaxies have been classified by visual inspection of their images. And in fact, technology is of little help here. It turns out that the human brain is far better than a computer at recognising the patterns that divide ellipticals from spirals. So visual inspection works well for a handful, or even several hundred objects....but one million? There are just too many galaxies for even the most dedicated of astronomers to look at. We need thousands of people to inspect galaxy images and to classify them as spiral or elliptical. We need you to help us.

You are given a quick primer on galaxy classification, then a test of your abilities to do so, and finally begin classifying galaxies. The really cool thing about this is that the galaxies (and a few other items) you see have not been seen by anybody before. Actually they probably run the same images by a number of viewers to see what percentage agrees on a particular classification.

The site makes use of the human eye's ability to classify images much too complex for a computer program to figure out but easy for us meat beings to do. Programs like Google Image Labeler use the same methods, but match you with a partner.

Check out "human computation" in Google Videos for an informative and enlightening lecture.

Human Computation

Hilton Ratcliffe · 08-10-2007

Hi Hill,

Thank you for posting this information. Astronomy is unusual in the applied physical sciences in having a significant "amateur" component, thousands of non-professional observers world-wide who use their eyes to assist professional astronomers with their work, and who make a significant, real contribution to progress in astronomy. Eminent British astronomer Sir Patrick Moore is one of them. While in no way mocking the very important message you send out here, I feel compelled to quote from my post yesterday in another thread. It refers to a paper submitted by participants at the recent Imperial College conference "Outstanding Challenges in Cosmology".

Quote:

But, heck, all is not smoke and thunder. In my ongoing study of the Cosmic Microwave Background, I came across a paper entitled “On the origin of the cosmic microwave background anisotropies”(arXiv:astro-ph/0703806). This publication probably says it better and more succinctly than any of the other thousand-odd that I’ve reviewed. All it requires is the ability to laugh at oneself. The first clue should have been the lead author, but I must admit I dived straight into the text without checking first. Who was it? Why, none other than one Ria Follop of the Institute of Fundamental and Outstanding Questions, Department of Cosmology and Metaphysics, Online University, Internet. Ahem!
The authors suggest that the background fluctuations may be correlated with topographical features on Earth, and use a statistical approach to make a powerful case for terrestrial, non-cosmic origins for the radiation picture. I will let them tell it in their own words: “To consider the signal that may be correlated with Earth, a correlate-by-eye exercise was attempted by overlaying the CMB map from Wilkinson Microwave Anisotropy Probe on a topographical map of Earth. Remarkably, several hot spots in the CMB map are found to be well aligned with either large cities on Earth or regions of high altitude. […] Simply extending our analysis, we suggest that cross-correlations between CMB and any other map of a Solar system body, image of a person, or an image of an animal will be detected at some statistical significance. […] Finally, we wish to comment on an existing suggestion in the literature that there are hidden messages from the Creator in the WMAP data since it can be thought of as a billboard visible throughout the Universe so a message is likely to be encoded within the intensity fluctuations. In the extreme case that the message is an image of the Creator herself hidden within the cosmic noise, we suggest that it may be possible to establish this image through a large number of cross-correlations of input images (of people, animals, spirits, or combinations) that are used in a likelihood analysis. […] In fact, we would not be surprised if a spherically projected image of a famous celebrity correlates with WMAP at some high significance. Thus it is left to the reader to establish using a complicated theoretical argument if that celebrity is the Creator whose image is then hidden in WMAP. […] We also thank the pub down Gloucester road for not closing at the time it should have closed to give us an extra half-hour to come up with the basic idea for this paper.”

Regards
Hilton

freeztar · 08-11-2007

Hello Hilton, thank you for adding to the forum in a beneficial manner.
If you would like to quote some text, please use the format of

Quote:

here you can see the code by quoting this

.

It is easier on the eyes for most.

coldcreation · 08-11-2007

Quote:

Originally Posted by freeztar