Earth Observation & GIS

GlobVision’s numerous contributions to the Earth Observation (EO) sector have been directed at application developments for the estimation of land surface parameters using remote sensing (e.g. satellite imagery) and in situ data. Our software applications can process a wide range of satellite imagery including high-resolution optical and all-weather synthetic aperture radar (SAR) imagery (single, dual and fully polarimetric).

Our software tools contain state-of-the-art feature extraction, pattern recognition, classification and machine learning algorithms developed in-house or by well-known researchers published in the literature. These algorithms convert optical and fully polarimetric SAR imagery to a set of informative features that are then used as input to the machine learning algorithms to infer desired ground parameters (e.g., wetland areas, land cover type, land usage, water extent, etc.). In addition, we have integrated a wide range of SAR backscattering inversion models that can estimate soil moisture and surface roughness from SAR backscattering coefficients.

In summary, over more than a decade and through several Earth Observation Application Development (EOADP) contracts with the Canadian Space Agency (CSA), we have developed a wide range of model-based and machine learning tools and algorithms for processing single-time and multi-temporal satellite imagery to estimate and monitor the change in ground parameters of interest to different end-users, industries, and environmental agencies.

For Geographical Information System (GIS), GlobVision has harnessed its expertise in artificial intelligence and data mining to devise architectures and learning methodologies to develop tools that complete and fill geographical information even in those areas where no sensors are available. This is the case, for instance, in seismic analysis as well as in meteorological monitoring.