Odd Radio Circle (ORC) are a circular object discovered by Australian Square Kilometre Array Pathfinder (ASKAP) while observing the Evolutionary Map of the Universe (EMU) survey. ORCs are anew unique class of astronomical object, in that they are only visible at radio wavelength. Currently there are less than 10 known ORCs discovered, and new discovery is difficult as amount of data produces by the telescope is too large to be visually inspected,

This project explores object detection algorithms such as Circle Hough Transform and Blob Detector as a method of detecting circular objects in radio astronomy images. To do this, images are first pre-processed to remove background noise. The different algorithms are run over the pre-processed image where the different search parameters such as sizes, shape, and edge detection are optimised for finding ORCS.
We test the two algorithms using images which contains three ORCs: Circle Hough Transform – which detected 2/3 ORCs, two Blob Detector algorithms where method one detected 3/3 ORCs, and method two detected 2/3 ORCs.

Some issue we’ve encountered include:
• Detecting ORCs which are too faint
• Detecting ORCs which could not be determined circular in shape

Future work should focus on exploring optimising preprocessing steps to pick up fainter images and more advanced machine learning algorithms such as Neural Networks to address some of the shortcomings in this project.