Finished! Looks like this project is out of data at the moment!
Hey everyone! thank you so much for all your hard work classifying the muon grids! We are back with more data - but this time, we are be back to classic Muon Hunters, classifying individual camera images and drawing circles through the muon rings.
Get classifying right now at: http://muonhunter.org/
FAQ
What am I being asked to do?
In Muon Hunters 2.0: Return of the Ring, we need you to determine whether our machine learning algorithm has done a good job of clustering our data. Our hope is that muon detections get clustered with other muons and that non-muons appear in separate clusters. We need your help to review how well this has worked.
How do I distinguish between Muon and Non-Muon detections?
The easiest way to spot a muon is that it will show a ring-like structure. If there are signs of a ring you should identify the image as a muon, even if a non-muon also appears in the same image.
Where can I see more examples?
To see more examples and for some explanation of why certain artifacts appear, check out the field guide to the right of your screen.
Why can't a computer do this?
This is one of the problems we are trying to solve and we are testing a new approach. The approach in the previous version of Muon Hunters was to gather classifications for individual images and then use that information to train a machine learning algorithm to identify muons and non-muons from the images. This time we are taking a different approach - here we let an algorithm group images it deems to be similar into clusters and then ask you for labels for each cluster. If the clustering is not very good then we can use that information to provide feedback to the algorithm to improve its performance for the future.
How is "muon" pronounced?
Muon is pronounced "MEW-on"
What do the colors mean and why do some images look different?
The color variation reflects the strength of the signal, from lowest to highest it goes blue->green->yellow->red->brown. This color scale dynamically adjusts to the strength of the brightest pixel to give us the maximum range to work with, making it easier to see images over the wide range of signal strength that we get, and enables us to see detail at all levels.
So for a bright shower this squashes the variation in the background down so we see less variation, making the shower seem sharper, but if the shower is weaker, the scale is lower so the background appears brighter.
Is this line actually an arc of a really big radius?
The muon ring/arc radius will always be smaller than the radius of the camera. This is due to the physics of the Cherenkov light emission process, which means the light in the muon ring does not go wider than about 2 degrees in diameter and the VERITAS camera is 3.5 degrees across.
How big a ring should I look for?
The ring size varies slightly depending on energy, but is dominated by the properties of the atmosphere, specifically the density, to have a radius larger than 1 degree. To give a sense of scale the camera itself is 3.5 degrees across, so any ring should have a large number of empty pixels within it. If there is only one empty pixel it is probably not a muon ring and more likely to be a pixel that was turned off, for example the photodetectors used are very sensitive to light and easily damaged so if there is a bright star in their field of view at that time a single pixel may be turned off.
Why do some rings look slightly elliptical?
It is a geometric effect. Look at a circle from its centre and it is a perfect circular shape; now move away from the centre, past the circumference, the further from the centre you are the more elliptical the circle will look. The same is true of the muon rings: when viewed from directly along the axis of travel of the muon they will be perfectly circular, but for most of the muons we are seeing we are not directly along the axis of travel so there will be a slight perceived ellipticity to them.