Photo plates have been used to capture and store astronomical images for quite a long time. In recent years, several projects are carried out to digitize photo plates and these digitized photo plates are shared through the Internet. We could extract invaluable astronomical data from digitized photo plates to analyse astronomical targets with very long temporal variations (up to decades). Extracting positions of celestial objects from photo plates and calculating their positions in celestial coordinates would be the first step. However, since astronomers would use multiple exposures to obtain images in photo plates and there are some scratches and mildews during storage of these photo plates, it becomes hard for us to directly obtain necessary information from digitized photo plates. In this paper, we will discuss the data processing pipeline developed by us to process photo plates digital archives.
Paper
Full text
Data processing pipeline for photo plates digital archives with deep neural networks
Semantic Scholar · Computer Science · 2022
Abstract
Photo plates have been used to capture and store astronomical images for quite a long time. In recent years, several projects are carried out to digitize photo plates and these digitized photo plates are shared through the Internet. We could extract invaluable astronomical data from digitized photo plates to analyse astronomical targets with very long temporal variations (up to decades). Extracting positions of celestial objects from photo plates and calculating their positions in celestial coordinates would be the first step. However, since astronomers would use multiple exposures to obtain images in photo plates and there are some scratches and mildews during storage of these photo plates, it becomes hard for us to directly obtain necessary information from digitized photo plates. In this paper, we will discuss the data processing pipeline developed by us to process photo plates digital archives.