SDT Satellite Ground Station and Image Processing

SDT proposes a wide range of customized products to her customers by harmonizing her image processing and satellite technologies experiences. SDT follows the needs of the users and related innovative methods in the world closely. 

She started its first works in the field of satellite technologies in 2010 with the ground station project of GÖKTÜRK-1 satellite. In the following years, he participated in and successfully completed ground station projects carried out with domestic and foreign partners.

In parallel, in the HASAT project, which is a satellite image evaluation project, she developed the module for the exploitation of SAR images and ensured its integration into different user environments. During the preliminary design phase of GÖKTÜRK-3 SAR Satellite, he carried out design activities for the system components on the basis of Sensor and Ground station solutions. SDT is ready to present all these experiences to the users in 2021 with ARTUK and GÖRGÜÇ products. The figure below shows the satellite and image processing projects in which SDT took place over the years, in time axis.

New Generation Ground Station Solutions

With its «Ground-as-a-Service» architecture developed as an alternative to traditional satellite earth stations; three basic segments of the ground system (namely Satellite Access (Access), Mission Control Ground Segment (MISSION) and User Location Segment (DATA)) provided that the need for new satellite/space market manufacturers to develop their own ground stations has disappeared and will provide serious optimization in terms of costs.

With the Ground-as-a-Service (GaaS) model, traditional satellite earth station information systems infrastructure has been transformed into service-based functions, and transformed into services that can be customized according to the needs of the clients.

Satellite provider, private-military institutions and organizations; The obligation to establish and maintain an integrated informatics infrastructure for their needs is eliminated with this approach and becomes a living system through scalable services.

Artuk

The main purpose of the ARTUK system is to provide all the necessary functions for users with the approach of the new generation satellite ground station as a service model using the following technologies;

Flexible query technologies (Semantic-NLP)

Micro service architectures

Virtualization, resource management and container infrastructure

Data transfer with block chain technologies

Artificial Intelligence, Machine Learning, Deep Learning

In this project, which is designed as a middleware application with a completely distributed architecture, new hardware resources can be easily added depending on the needs and can work in harmony with existing resources.

In terms of satellite data providers, it ensures that the image products they sell become more value-added with the smart evaluation methods triggered in line with user demands or according to the data type.

Image Processing and Deep Learning Capabilities

SDT has been involved in many projects on the analysis of data/images from different sensors, besides, she has significantly increased the variety of sensors and algorithms with the R&D projects she has developed with its own resources. Namely, it has added EO / IR image processing, Lidar point cloud data processing, Hyperspectral image processing capabilities to her SAR image processing capabilities, which she has specialized over the years, and has also worked on multi-sensor data fusion in a specific area. In the image processing competition, which was opened with the cooperation of SSB-OGAM in 2020, SDT was ranked first in recognition of two targets and the third in average.

Within the scope of the project “Forest data analysis with Lidar” which is supported by TUBİTAK, the team has collected data with a small-type UAV and integrated Lidar system and developed algorithms for forest land analysis.

SDT image processing team has also developed algorithms for embedded target identification by processing hyperspectral data, and submitted it to the user authority.

SDT has been developing the target recognition algorithms based on deep learning since 2019, by using the rich image databases she has acquired over the years. She has achieved quite successful results in these studies where the number of labeled data and hardware infrastructure is critical.

SDT has just released the GÖRGÜÇ desktop software, which offers users all the image processing capabilities together with a toolbox architecture. GÖRGÜÇ software will continue to develop every year and become a living project with the feedbacks from users, and will be combined with our cloud-based ground station solutions in 2022 and presented to the user with a service model. With these approaches, which are very new in the world, the user will be able to use many algorithms and functions as needed without having to make necessary hardware and software investments.

GÖRGÜÇ

GÖRGÜÇ Toolboxes

Basic functions and preprocessing: GIS functions, image visualization, zoom in/zoom out, measurements, vector processes etc)

Pre-processing (median filter, histogram equalization, speckle noise reduction, gamma filter etc)

SAR Toolboxes

SAR -1: Image Analysis toolbox (land cover classification, road/railroad detection, change detection, shadow analysis, target detection, similar regions)

SAR -2: Target recognition (airplane, ship, bridge, runway)

SAR -3: Enhanced target recognition (POL, helicopter, tank)

SAR Interferometri Tool boxes

SAR Interferometri-1 CCD (Coherent Change Detection)

SAR Interferometri-2 DEM (Digital Elevation Map)

EO Toolboxes

EO -1: Image analysis toolbox (land cover classification, change detection, similar regions)

EO -2: Target recognition (airplane, ship, bridge, runway)

EO -3: Enhanced target recognition (building, POL, helipad, tunnel)

Supported SAR image types:

SARPER - National SAR sensor: Stanag-7023 format

TerraSAR-X X band SAR Satellite: COSAR format

Sentinel C Band SAR Satellite: Sentinel SAFE format

Cosmo Sky Med X band SAR Satellite: H5 format

Supported EO image types:

GÖKTÜRK

Sentinel -1