DECO: Image Analysis

As we have discussed in our talks and other blog posts, noise is a large factor in creating events, and we don’t want noise, whether it be thermal, the most common, or magnetic. One of the great things about being able to actually see event images on a computer is that an event created by noise is easy to identify based off the pixels seen in the image. For example, events created by noise rarely have a small group of pixels at a higher brightness or intensity in comparison to the other pixels in the image. Events created by cosmic rays have a grouping of pixels that is significantly brighter than the surrounding pixels. We can see these clusters of pixels by altering the color range of the image. Because the Charged-couple devices (CCDs) in the cameras only use red, green(2), and blue, all we have to do is distort the color range so that those colors are more pronounced. If the color range of an image is reduced to all green, it is easier to spot the bright spot(s) in the image. If you don’t reduce the color range, the image is black and the brightened pixels are almost impossible to see. Now it’s just a process of filtering out events created by noise.

Over the past few months, many queries have been conducted and many numbers recorded in regards to pixel intensity. A simple way to eliminate events created by noise is to come up with a set minimum pixel intensity per image. While this method may not eliminate all events created by noise–and very well could eliminate lower energy cosmic ray particles–it will significantly reduce the volume of events created which makes analysis much easier. There is also a higher certainty that the events registered are cosmic rays because the filter will only show high-energy events. Determining the minimum pixel intensity is the current focus of the project on a broader scale.

In terms of the tests run in Wyoming to determine if there was a relationship between altitude and flux, please visit this section of our wiki page.

Leave a Reply