I am attempting to write code to analyse .FITS images.
My goal is to find those images that contain moving objects (asteroids), which appear as streaks (lines). I would like to reject images that contain just stars (points of light).
There is a rather large variance in image background, and streak size, that I am dealing with.
I am very new to coding, so would appreciate any suggestions to achieve my goal. After reading through examples and documentation online, I have tried thresholding using 'graythresh', then 'bwmorph' processes before the 'houghpeaks' process to detect lines.
I am not committed to any one way of line detection – any suggestions whatsoever are welcome and much appreciated!
Many thanks in advance.
Best Answer