The research project LEO-SITS aims at enhancing the functionalities of the EO Payload Ground Segment systems and the next-generation services for an enhanced and long term data expatiation of the ITS already acquired and to be continued by the Sentinels, with focus on Sentinel 2.
The need to reach historical Earth Observation (EO) data series strongly increased in the last years, mainly for environmental monitoring applications. This trend is likely to increase even more in the future in particular for the growing interest on global change monitoring that requires data time-series spanning 20 years and more. Considering new sensors spatial resolution giving access to detailed image structures, the opportunities to compose high-resolution satellite image time-series (SITS) are growing and the observation of precise spatio-temporal structures in dynamic scenes is getting more and more accessible.
Dedicated tools for information extraction in SITS have been developed in order to perform change detection, monitoring, or validation of physical models. However, these techniques usually are dedicated to specific applications. Consequently, in order to exploit the information contained in SITS, general analytical methods are required.

This project is focused on information extraction in the form of categories of evolution. Project goal is to develop algorithms and techniques to classify the evolutions of several categories inside a scene. It is of great importance for a user searching in SITS archives to find and delimitate classes having the same evolution in time.
The main idea is to model an image time series based on the transformations occurring between consecutive acquisition of the considered SITS. The novelty of the proposed technique lays in the conversion of the multispectral information from the entire data set into a change map time series (CMTS).
Pairs of consecutive images will be described by a number of change maps computed using similarity measures. Complementary information about the changes occurred in the scene will be extracted and further employed in order to provide the user with a broader perspective on the land transformation processes.

Project page: http://leosits.ceospacetech.pub.ro