Objects recognition on the astronomical images

sobota 2024‐04‐20 10:00 – 11:00
  • Fakulta matematiky, fyziky a informatiky, Univerzita Komenského v Bratislave
  • Astros Solutions s. r. o.

The astronomical Image processing pipelines are were crucial and complex systems. Nominally, the pipeline starts with the image calibration, subtraction of the background or calibration frames, then comes the segmentation and detection of the frame objects on images. Once all objects are detected, correlation with astrometric and photometric catalogue can be done to estimate the conversion to the standard coordinate systems. At this point comes the topic of the practical exercise i.e., correlation of the frame object corresponding to the object itself into the series of angular positions.

The general aim of the exercise is call for ideas for the object correlation algorithms. We will prepare the testing series of images, which will be segmented, astrometrized and masked for background stars. This series, along with the basic data management tools will be provided to the audience and task for the exercise will be to define the algorithm, which can be used to form the series of angular positions, containing only the target’s position on each image. If also some prototype can be established, it will be considered as beneficial, but aim is to formulate the algorithm.