DeepSpaceYoloDataset: Annotated Astronomical Images Captured with Smart Telescopes

Abstract

Recent smart telescopes allow the automatic collection of a large quantity of data for specific portions of the night sky—with the goal of capturing images of deep sky objects (nebula, galaxies, globular clusters). Nevertheless, human verification is still required afterwards to check whether celestial targets are effectively visible in the images produced by these instruments. Depending on the magnitude of deep sky objects, the observation conditions and the cumulative time of data acquisition, it is possible that only stars are present in the images. In addition, unfavorable external conditions (light pollution, bright moon, etc.) can make capture difficult. In this paper, we describe DeepSpaceYoloDataset, a set of 4696 RGB astronomical images captured by two smart telescopes and annotated with the positions of deep sky objects that are effectively in the images. This dataset can be used to train detection models on this type of image, enabling the better control of the duration of capture sessions, but also to detect unexpected celestial events such as supernova.

Deep Sky Objects Detection with Deep Learning for Electronically Assisted Astronomy

“…As described in [28,29], collecting astronomical images from an area heavily impacted by light pollution, such as Luxembourg, France, and Belgium, posed a significant challenge due to the need for a long-term observation period. EAA is an outdoor activity that is subject to weather conditions, requiring us to be readily available when suitable context arose.…”

“…On the one hand, raw data captured with Stellina corresponds to 205 observation sessions, 50068 FITS images of resolution 3096 × 2080 were obtained (with a field of view of approximately 1 × 0.7) [28]. On the other hand, post-processed images obtained with both Stellina and Vespera are published as a set of 4696 astronomical RGB images in JPEG format (minimal compression) with a resolution of 608 × 608 pixels [29].…”

“…As a results, we obtained DeepSpaceYoloDataset [29], a dataset formatted with the YOLO standard, i.e., a ZIP file containing 4696 RGB images in JPEG format (minimal compression), and 4696 text files containing the positions of DSO, with separated files for images and annotations, usable by state-of-the-art training tools [35].…”