Thanks
I started in 2012 and it took me a year before I had my first comet discovery, a Kreutz.
It's not hard to find a comet, there's a new one along everyone couple of days but the competition to be the first to report that comet is fierce. Some people even automatically download the raw FITS images as they come in and process themselves rather than wait 10 minutes for the processed images to be made available.
In light of this it wasn't a good use of my time to continue. However in 2013 I had studied a unit in Python at MIT EDX. For a few years I always had this idea of writing a python program to search through the archival images to look for missed comets. in 2019 I wrote my first draft of a program to do that and discovered a number of comets in the archives. However early versions of the program were slow and the output of the program required me, for each candidate found, to go back and look at the original images to judge whether what had been found was real or not. Using a couple of python libraries I have since been able to speed up my program significantly and also change the output to a series of postage stamp images that enable me to instantly tell me what has been found is a comet or not. For example:
Not a comet
Comet
Even so it takes quite a while to search through the candidates found. I am digging deeper and deeper into the noise and as one does so the number of false positives rises markedly.
This possibly means I would be the 2nd highest comet discoverer in Australia currently. Robert McNaught is the leader with 82 I believe.
I also have another couple of comet discovery announcements to make in the near future but not with soho images. I am just waiting on confirmation of the discoveries.
cheers Peter