A general purpose automated masking algorithm for deconvolution was developed in order to support automated data processing in ever-increasing data volumes of the current and future radio interferometers as described by Kepley et al. in this meeting. In this presentation, we describe some technical details of the implementation of the automated masking algorithm named “auto-multithresh” which was integrated into the refactored imaging task (tclean) in CASA. We also discuss our approach that we took for the development, which loosely follows the iterative model, so that the implementation is refined progressively for its functionality and performance based on testing and updated requirements throughout prototyping in Python to the final production in C++.
Link to PDF (may not be available yet): P12-15.pdf