Marc Masias (Universitat de Girona),
Jordi Freixenet (Universitat de Girona),
Xavier Llado (Universitat de Girona), Marta Peracaula (Universitat de
Girona)
Abstract
In astronomical imaging great efforts have been
done aiming to perform automatic detections of stellar bodies in wide
field images or in large surveys. These images contain large volumes
of data at different intensity scales and often have a high component
of noise. For these reasons, even the source detection done by visual
inspection is a challenging task. With the idea to improve this source
detection process, different automatic tools have been proposed. We
present in this paper a review of the main approaches for the
automatic detection of sources in astronomical images. Several
approaches that use images at different frequency bands are analysed
in order to find out their advantages and drawbacks, and to specify
the key points that make them remarkable for the astronomical
community. We classify the most important techniques into different
strategies according to the type of pre-processing applied and the
strategy used to deal with the detection problem. A qualitative and
quantitative evaluation of the results of the most outstanding
approaches is also presented. Finally, we discuss on the open research
fields of the astronomical object detection.Paper ID: P094
Poster Instructions
|
