Remote sensing and environment information systems
We develop advanced image interpretation methods to retrieve information from
digital satellite images. Our special fields of expertise are remote sensing
of the natural environment, particularly forestry applications and sea ice
monitoring. Other important focus areas include security applications and
monitoring methods for sea and terrestrial traffic. Technologically the main
interest presently is in the analysis of multi-source data, very high
resolution data, and imaging radar data. Automatic analysis of ground-based
digital photographs to support satellite image analysis is a new extension of
the research interest.
Time-series of ALOS Palsar radar satellite images from Penang, Malaysia. Changes can be seen in the coloured areas, mainly in the field areas due to the growing of grain. ©JAXA and METI 2007.
In general, interest and investment in remote sensing are increasing. The increased interest is due to the concern with the sustainable use of natural resources globally and the climate change issue. High costs of personnel in industrialised countries and the poor physical accessibility of many regions are boosting the use of earth observation techniques. Another important driver is preparation for natural disasters and mitigation of the damage.
Presently we are developing solutions for, among other things, forest inventory to support the REDD (Reduction of Emissions from Deforestation and forest Degradation) process, including automatic tree-wise analysis of high resolution satellite images. Another example is the combination of ship and terrestrial radar imagery with satellite radar images to improve the safety of winter marine traffic and decrease fuel costs.
The environment information systems integrate tools and methods into distributed services.
Reduced costs (e.g. wood procurement, fuel consumption in winter sea traffic)
Earlier information (e.g. increased flood risk)
Unique information (e.g. global forest resources)
Global and local (image resolutions from 0.5 metres to 1 km)
Improved safety (e.g. safe routes for ships)
Harmonised and verified information quality (similar data at all locations)
Flexibility (all data with coordinates can be assimilated)
Historical data available in original form for enhanced analysis (i.e. images)
References and merits
Key partner in international projects including the GMES Service Element projects Forest Monitoring of ESA and Geoland2 in the European Commission’s Seventh Framework Programme.
Probability method and software for the estimation of forest biomass and other variables. In operative use in locations worldwide.
Image segmentation method Pyramidas.
Pre-processing software and algorithmic tools for the new generation satellite
IBNet: operative distributed traffic information system for icebreakers.
ShipSensornet: satellite radar, terrestrial and ship radar images are merged
to improve the quality of ice monitoring in winter navigation.
EnsoMosaic Aerial Digital Imaging and image mosaic compilation. Outsourced
Tree species classification from very high resolution satellite imagery. (red = deciduous, green = spruce, blue = pine)