Development of automated methods for processing and analyzing images from space missions

Project realized by:

Jagiellonian University, Faculty FAIS

Aim of the project:

Developing tools for processing, analyzing, and visualizing data from space missions.

Project description:

Due to the large number of images being received from space missions, it becomes necessary to process them automatically rather than manually. Using data processing and analysis methods, including machine learning techniques, we develop dedicated automatic and semi-automatic tools for:

  • Image segmentation and their subsequent analysis
  • Classification of objects – e.g. recognition of forms, and examining their parameters, such as orientation, width, length etc.
  • Rectification of images, aligning images together, including overlapping images of the same area over several different periods of time
  • Object motion tracking – checking whether the object visible in a series of images, taken at intervals, moves/changes
  • Calculation of landforms’ volume based on DTM and their derivatives
  • Digital visualizations of studied processes (animations, simulations)

Project methodology:

The software used in this project: ISIS, Matlab, Mathematica, Godot Engine, Unreal Engine, ArcGIS, Adobe Creative Cloud, and developed tools.