NaturaSat

The NaturaSat application combines geobotanical research with modern mathematical methods. It utilizes data from the Sentinel-2 satellite, orthophotos, and UAVs to identify precise boundaries of habitats, enhancing knowledge about their current distribution and spatio-temporal changes.

The software integrates various image processing techniques along with vegetation data into one universal tool, allowing for comprehensive exploration of habitats. In addition to segmentation, it also offers other tools such as monitoring, classification, transformation of historical maps, and basic hydrological modeling.

Tools available in the Trial version of the application

The NaturaSat program contains a wide range of tools that enable the user to perform full segmentation of habitats and their subsequent monitoring. In the trial version of the application, the user can try out basic operations with the program for up to 30 days. The user has the opportunity to download, load, and display satellite images from the Sentinel-2 mission, use semi-automatic and automatic segmentation tools, and save resulting polygons in KML format. Statistical changes on the created polygons can be monitored using the habitat monitoring tool.

Displaying Sentinel-2 mission images

The tool allows the user to download, load, and display satellite images obtained from the Sentinel-2 mission.

Semi-automatic segmentation

Semi-automatic segmentation enables the segmentation of areas using nonlinear segments created by the user with a mouse.

Automatic segmentation

After entering parameters, the tool allows automatic segmentation using evolution from the initial polygon towards the boundaries of areas.

Habitat
monitoring

The monitoring tool enables tracking statistical changes computed from optical channels of satellite images from different periods.

Paid tools of the application

Within the paid version, the user unlocks the full potential of the NaturaSat application. This allows the user not only to load data from the Sentinel-2 mission but also orthophotos and UAV images. In addition to segmentation and monitoring tools, the paid version unlocks tools for classification, transformation of historical maps, and basic hydrological modeling, enabling the user to conduct even more detailed analysis of the studied habitats.

Support for multiple input data

The application supports loading orthophotos, UAV data and other standard image formats.

Polygon manager

The polygon manager allows users to manage created and loaded segmentation polygons.

Classification

The classification tool enables users to classify areas using deep learning techniques.

Transformation of historical maps

The tool utilizes both local and global transformation methods to find corresponding areas on historical maps and the current interactive map.

Basic hydrological modeling

Hydrological modeling enables the calculation of the approximate height of groundwater and surface water table in river system areas.

Paid services

Training of the NatNet Networks

The classification tool is based on deep learning methods and utilizes Natural Numerical Networks (NatNets).
If interested, we offer network tuning for a selected group of habitats, with consultations provided by users during the preparation of the training dataset.

Downloading and displaying satellite images

The NaturaSat application allows users to load and display all optical channels of satellite images obtained from the Sentinel-2 mission. New images can be downloaded directly through the user interface using the Copernicus Data Space Ecosystem API.

Semi-automatic segmentation

The semi-automatic segmentation tool allows users to input boundary points of an area using the mouse. These points are then connected to form nonlinear segments of a polygon, which replicates the boundaries of the segmented area. After completing the semi-automatic segmentation or closing the polygon, the resulting polygon is saved in the user interface’s list, from where it can be exported for further use.

Automatic segmentation

After entering parameters, the tool allows initiating the evolution of the initial polygon. This polygon is evolved by a mathematical model of automatic segmentation towards the boundary of the selected area. The resulting polygon and its positions during evolution are saved in the user interface’s list, from where they can be exported for further use.

Monitoring of habitats

The habitat monitoring tool enables the calculation of statistics from the intensities of optical channels of satellite images located within selected segmentation polygons. Statistics computed from different time periods can be compared, allowing users to monitor significant changes over time.

Classification

The classification tool is based on deep learning methods and utilizes Natural Numerical Networks (NatNets). The result of the classification is relevance maps, which indicate the potential occurrence of classified habitats in the input image.