Photometry pipeline
During the summer of 2023 I worked alongside Juan Santisteban of the University of St Andrews to develop a source identification and photometry pipeline. The goal was to feed this pipeline a folder of astronomical .fits images, have it identify all astronomical sources, then produce a series of light curves from selected targets.
The most significant challenges in this project stemmed from target tracking and ensuring that the code would be able to reliably identify the same source regardless of image transformation or rotation.
This pipeline will be used to characterise the OPTICAM imaging system, then produce and analyse photometric data from selected cataclysmic binary systems.
current status
The pipeline can now load all fits images from a specified directory, identify sources within those images, and perform photometry on them automatically. Most importantly, it can assign IDs to each star so that they can be tracked across each image. This was achieved using the AstroAlign library and by storing sequences of coordinates corresponding to the positions of each star across the different images. As of now AstroAlign itself is used to detect sources, but the issue we're currently facing is that if it fails to detect the transform between two images then we cannot assign the correct ID to stars in subsequent images.It seems to be that if the list of coordinates passed into the find_transform function is too short, AA exhausts the possible configurations in which those sources could be matched to those in the subsequent image before finding the correct transform.
This is the first successful lightcurve output from a series of one thousand images.
