WEBMAIL    |    Intranet    |    Site Map





Coordinator: ROSA - Romanian Space Agency


CRUTA - Romanian Center for Remote Sensing in Agriculture
National Research Institute for Land Reclamation - INCDIF-ISPIF
ASRC - Advanced Studies and Research Center
"Polytechnic" University of Bucharest

Period: 36 months (2007 to 2010)

Project director:

Description: In recent decades, the Earth observation domain has been supported by numerous SAR space missions, such as: EERS1/2, Envisat, Radarsat, SRTM who took huge amounts of data. Thanks to radar systems' capacity to operate independently of weather conditions, day and night, and also to the coherent nature of data, there were developed many applications of great importance: mapping, monitoring hazard / crisis, DEM generation, evaluation of landslides, assessment of seismic movements, observations of oceans and coastal areas, etc. However, SAR data processing and information retrieval are not easy tasks, requiring multidisciplinary techniques from the fields of radar technology, electronics, signal processing, statistics, physics phenomena of dispersion, geospatial sciences, data and models assimilation, geodesy and so on.

There were recently released on the orbit new space missions with high resolution SAR sensors (0.5-1 meter): the German satellite TerraSAR and Cosmos Skymed were launched in June 2007, while RADARSAT-2 is planned for the second part this year. The interpretation of with the 1 meter resolution SAR data will be challenging and complex. Therefore, this project aims at the development and integration of support tools for applications using high resolution SAR information and to demonstrate some relevant applications.

Synthetic aperture radar systems (SAR) use coherent imagery and processing techniques to generate high resolution images of the land, regardless of time and weather conditions. The high resolution SAR images the structures can be observed both in amplitude and phase. The images taken by SAR are different from the optical ones, the interpretation of which is a difficult process because the information recorded in SAR image formation and acquisition is different from the visual perception. The contents of a SAR image is characterized by its own geometry, which is different from the actual geometry of a lightened scene and is dominated by strong dispersion phenomena.

Based on these principles, the purpose of RDI activities will be to investigate the ability of new sensors to solve the general problems applicalble to the local and regional mapping process within the GMES Land Monitoring Core Service, such as improving the separation of classes cover / land use in the automatic interpretation, identification of urban structures, monitoring construction areas, identifying class land use change, soil mapping, monitoring of specific objectives. The results will be tested taking into account specific issues such as: identifying isolated buildings, detecting forest boundary line, delimitation general object recognition tree crown detection and evaluation of deforested areas within the forest, identify vertical structures developed.

The research will result in the development of integrated tools in the processing flow for executing certain processes: change detection, interferometry, data fusion (SAR data, optical and auxiliary data); classification and image analysis using methods based on objects.

Project objectives: To identify the needs and requirements for the use of SAR data and applications - establishing contacts with national users working in various fields that can benefit from SAR observations and can gather requirements and define and summarize scenarios of relevant case studies.

Creating a center of expertise in SAR - organizing a SAR processing chain to solve the processing needs for the most important applications. Providing training in key areas of theory, technology and applications of SAR. Creating a SAR compenence center, a consortium of multidisciplinary teams that aim to act as a national focal points and points of contact for expertise, knowledge and information, and to support all operations and SAR applications.
Establishment of new mechanisms for the participation of private sector organizations in the support the further development of initiatives related to the project - the aim is to achieve a network of organizations working together.


Step 1 - Analysis of SAR systems, features, products and areas of use and the development of a reference database composed of SAR images 

Step 2 - Research and implementation of algorithms for extracting descriptors of complex single SAR images and detection methods for changes in non-coherent sequences of SAR images

Step 3 - Methods of change detection for coherent sequences of SAR images and tools for machine learning and pattern recognition

Step 4 - Evaluation and validation of algorithms, methods and tools and studies demonstrations on real cases 

Step 5 - Organizing the existing tools, developing new modules and future prospects analysis of methods developed for use in InSAR high-resolution data

Exploitation and dissemination measures are the main activities of the proposal, whose relevance can be seen early in the RO-SAR project. The whole consortium will conduct these activities under the supervision of the Project Coordinator - the Romania Space Agency and the leaders of the activities.

Exploitation and dissemination activities will continue after the completion of the proposed project. The topic of the disseminations will be:

  • general information on the RO-SAR project and its activities
  • materials (guides, checklists) that will be sen to targeted groups (research community)
  • training materials (developed at a hogh level and distributed at local/regional/international level
  • information given during the national and international training sessions
  • conferences and training sessions reports


Project website: RO-SAR