Quick, Look! A New Method for Artifact Detection

Abstract

In the era of all-sky surveys, wide-area radio surveys like the Very Large Array Sky Survey (VLASS) place strong emphasis on the identification and classification of candidate transient events. To aid the transient searches, automatically generated VLASS Quick-Look (QL) images are designed and produced to deal with high data volumes and enable rapid followup observations. But, the incomplete and snapshot sampling of the sky in the uv-plane during VLASS observations combined with imaging choices made when generating QL images often lead to residual linear artifacts, particularly around brighter sources. While well-established techniques (like clean) maximizes the information from this partial sampling, such techniques can still be imperfect. Therefore, the need for automatic image-quality classification and assurance is more important than ever to ensure rapid data quality assessment and for enabling the best science. In this thesis, I present a new technique to identify these linear streaks around sources detected in the VLASS Epoch 1 QL images by extending the results of a line detection technique called the Hough Transform. After robustly quantifying the identified streaks, their effects on the sources/components that they are overlapping are removed. The resulting artifacts-subtracted components’ bright- nesses are then used to distinguish real astrophysical sources from imaging artifacts. Finally, I discuss the use of this streak detection method as an additional quality assessment step during interferometric image reconstruction.