Integrated Analysis of Satellite Imagery for Nuclear Monitoring - Results from G-SEXTANT

The FP7 project G-SEXTANT (Geospatial Intelligence Services in Support of EU External Action, 2013-2014) intends to develop a portfolio of Earth Observation products and services to support the geo-spatial information needs of EU External Action users and stakeholders. G-SEXTANT is part of Copernicus, a European Earth Observation programme which combines the use of satellite images and data with local, in situ, data sources to deliver geo-spatial information services and products to a wide range of end-users. Copernicus is expected to create an autonomous and operational European capability in environmental and Security information services. G-SEXTANT contains a work package entitled “Nuclear activities scenario”, which provides tools in support of monitoring nuclear-related sites and activities using satellite imagery. The activities of the work package are grouped into two sub-scenarios, namely “Monitoring of nuclear decommissioning activities” and “Monitoring of nuclear activities in the context of the Nuclear Non-Proliferation Treaty (NPT)”. The paper presents the developments of the second sub-scenario, in detail: i) SAR Time Series Analysis using the Method for generalized Means Ordered Series Analysis (MIMOSA), including the following steps: estimation of the amplitude distribution, computation of the probability density, and automatic thresholding. The procedure is computationally efficient without need for spatial speckle filtering. ii) SAR Change Detection using the Complex Wishart Algorithm with dual and quad polarimetric imagery in look-averaged covariance matrix format in order to define a per-pixel change/no-change hypothesis test. The method includes approximations for the probability distribution of the test statistic, and so permits quantitative significance levels to be quoted for change pixels. The processing chain generates geo-coded change maps at the desired statistical significance level. iii) Optical Change Detection using Object-based Techniques. Starting with the segmentation of the images of both acquisition times, the process transforms the feature space using multivariate statistical methods in order to emphasize and classify relevant changes. iv) GIS-based integration of the products in a system which provides access to all available products and information via standard web interfaces. The system is able to integrate and manage multi-temporal and multi-type information (satellite imagery, Open-Source documents, 3D-data information, GIS data).

NIEMEYER Irmgard;  LISTNER C.;  CANTY Mort;  WOLFART Erik;  LAGRANGE Jean-Michel; 

2014-12-05

Institute for Nuclear Materials Management

JRC90528

https://publications.jrc.ec.europa.eu/repository/handle/JRC90528,