Satellite and astronomical images processing and analysis
Super-resolution reconstruction of satellite images
We develop algorithms that allow us to generate images with a higher spatial resolution based on a single- or multiple observations captured for the analyzed scene. The acquisition of lower-resolution images is much faster and cheaper in practice, and it is often infeasible to obtain and efficiently transfer high-resolution images, e.g., from an imaging satellite. Super-resolution reconstruction may be considered an initial step in the processing chain in Earth observation – increasing the spatial quality of input images may significantly improve the performance of the further steps of the processing pipeline, such as object detection or image segmentation. Our techniques can be exploited not only in satellite imaging, but are applicable for super-resolution of color images and video sequences.
Astronomical image analysis
Image acquisition techniques are key in astronomy, as image data is commonly our exclusive source of information about the characteristics of observed objects. Our algorithms include the approaches for improving the quality of such data through various corrections, alongside the optimization techniques that help improve the acquisition process. We also focus on the analysis of astronomical images using classical machine vision and image processing, together with machine learning.
Contact (email: firstname.lastname@polsl.pl):
- Dr hab. inż. Adam Popowicz, Prof. PŚ (ORCID)
- Dr inż. Krzysztof Bernacki
