|
Intelligent Systems for Environmental
Monitoring and Sustainable Development
Overview
- Computing science research for environmental monitoring systems.
- Complex, interdisciplinary projects involving Canadian and
international satellites and research teams.
- Canada: radar satellites-Radarsat-1; Radarsat-2 (~2005);
hyperspectral satellite-HEROS (~2008).
- International: US-Landsat, EO-1, Terra, Aqua; ESA-ENVISAT, PROBA.
- Commercial: 0.5 m up (~10 satellites).
Objective and Activities
- Primary objective is to develop intelligent computer vision and data fusion methods and systems to
dynamically select sensors and features to achieve the highest recognition accuracies at the lowest costs.
- Activities: hyperspectral analysis methods for determining spectral endmembers; creation of methods
and system for creating Kyoto Protocol information products; high-resolution multiangle analysis for
determination of 3-D forest properties; expert systems for data fusion; distributed data analysis
over high bandwidth links with GRID architectures.
Project Examples
- Automatic system for hyperspectral remote sensing endmember unmixing
- Kyoto Protocol Advanced Classification System
- An intelligent system for measuring from space the chemical properties of Canada's forests
- A system to analyze 3-D multiangle imagery for estimating the 3-D properties of forests
- Distributed hyperspectral and segmentation processing over GRID architectures
SAFORAH
- Distributed grid computing for the storage and processing of large remote sensing data sets
- Development through collaboration of government, university and industry
- Facilitates national forest monitoring and research activities
|
