Eliminating artefacts in Polarimetric Images using Deep Learning

TL;DR

This paper presents a deep learning approach using CNNs to automatically detect and classify artefacts in polarimetric images, improving image quality for wide-field surveys.

Contribution

It introduces a CNN-based method for artefact detection in polarimetric images, achieving high accuracy and facilitating automated quality control in large surveys.

Findings

Achieved 98% true positive rate in artefact classification

Achieved 97% true negative rate in artefact classification

Demonstrated potential for integration with transfer learning for future surveys

Abstract

Polarization measurements done using Imaging Polarimeters such as the Robotic Polarimeter are very sensitive to the presence of artefacts in images. Artefacts can range from internal reflections in a telescope to satellite trails that could contaminate an area of interest in the image. With the advent of wide-field polarimetry surveys, it is imperative to develop methods that automatically flag artefacts in images. In this paper, we implement a Convolutional Neural Network to identify the most dominant artefacts in the images. We find that our model can successfully classify sources with 98\% true positive and 97\% true negative rates. Such models, combined with transfer learning, will give us a running start in artefact elimination for near-future surveys like WALOP.