# A Computer Vision Approach To Identify Einstein Rings And Arcs

## TL;DR

This paper introduces a computer vision method using the circle Hough transform to automatically identify Einstein rings in large astronomical survey data, facilitating the discovery of these rare gravitational lensing phenomena.

## Contribution

The work presents a novel automated approach combining pre-selection of lens candidates with circle detection, enabling efficient identification of Einstein rings in big data surveys.

## Key findings

- High completeness in detecting Einstein rings
- Successfully applied to multiple large sky surveys
- Operates solely on JPEG images without pre-processing

## Abstract

Einstein rings are rare gem of the strong lensing phenomena. Unlike doubly or quadruply lensed systems, the ring images can be used to probe the underlying lens gravitational potential at every position angle, putting much tighter constraints on the lens mass profile. In addition, the magnified background source also enable us to probe high-z galaxies with enhanced spatial resolution and higher S/N, which is otherwise not possible for un-lensed galaxy studies. Despite their usefulness, only a handful of Einstein rings have been reported so far, mainly by serendipitous discoveries or visual inspections of hundred thousands of massive galaxies or galaxy clusters. With the on-going and forth-coming large area surveys such as Large Synoptic Survey Telescope, visual inspection to discover Einstein rings is very difficult, and an automated approach to identify ring pattern in the big data to come is in high demand. Here we present an Einstein ring recognition approach based on computer vision techniques. The workhorse is the circle Hough transform, which can recognize circular patterns or arcs at any position with any radius in the images. We devise a two-tier approach: first pre-select LRGs associated with multiple blue objects as possible lens galaxies, then feed these possible lenses to Hough transform. As a proof-of-concept, we investigate our approach using the Sloan Digital Sky Surveys. Our results show high completeness, albeit low purity. We also apply our approach to three newly discovered Einstein rings/arcs, in the DES, HSC-SSP, and UltraVISTA survey, illustrating the versatility of our approach to on-going and up-coming large sky surveys in general. The beauty of our approach is that it is solely based on JPEG images, which can be easily obtained in batch mode from SDSS finding chart tools, without any pre-processing of the image. (Abridged)