Hypermix
Description
|
HyperMix |
| Advanced airborne hyperspectral remote sensing to support forest management |
| Contact |
Stephanie DELALIEUX
Birgen HAEST |
Start (End) Date |
01/12/2010
(01/12/2013) |
| Consortium |
1 partner |
Project Coordination |
VITO |
| Website |
- |
Related Projects |
Habistat, MsMonina,
DigiKart, HyperKart |
| Keywords |
hyperspectral, hyperspatial, information and data fusion,
spectral-spatial trade-off |
Project Objectives
Remote sensing has proven to be a powerful tool for the monitoring of the Earth’s surface and atmosphere on a global and local scale. However, sensor limitations are most often a serious drawback since no single sensor combines the optimal spectral, spatial and temporal resolutions. Currently there are only a few very high spatial resolution (<1m) sensors all having a limited number of spectral bands (e.g. Ikonos and Quickbird) and there are only a few very high spectral resolution sensors with a spatial resolution of 18-30m (e.g. Hyperion and Chris). Images are however acquired upon user request and have a limited spatial coverage. With the launch of new very high resolution satellites like Worldview -2 and planned hyperspectral missions like Enmap, Prisma and Hyspiri much more data will become available to the user community. Still, the trade-off in spectral and spatial resolution will remain and new advanced data and information fusion approaches are needed to make optimal use of these future sensors. Therefore, in this project, techniques will be developed for the fusion of hyperspectral images with images of high spatial resolution. Since most fusion techniques are limited to the fusion of multispectral images with panchromatic images, we will focus on the extension of such techniques towards hyperspectral images.
Methodology
In order not to be tied to a specific satellite sensor, this project will start the fusion process with the most detailed information available both spectrally and spatially. Hyperspatial (cm) images will be gathered by highly flexible micro unmanned aerial vehicles (micro-UAV’s), hyperspectral data will be acquired with the APEX sensor. These datasets provide a perfect starting point from which minimum spatial and spectral requirements can be specified. The Kalmthoutse Heide has been selected as study area since remote sensing studies on vegetation monitoring and phenological research in such heterogeneous landscapes with gradual transitions of vegetation patches will certainly derive benefit from high spectral and high spatial images. Moreover, the classification framework developed within the Habistat project on the application of habitat status reporting in which both partners are involved, will be applied to perform a validation exercise.
(Expected) Results
New methodologies for the information fusion of hyperspectral and hyperspatial data, published in peer-reviewed journals and presented at international conferences. A requirement analysis for future spaceborne sensors will also be performed..
VITO Contribution
VITO is the coordinator of the project and is involved in the algorithm development on information fusion.
Partners
University of Antwerp (Belgium)
Contact:
Stephanie Delalieux
Tel. + 32 14 33 67 16
Fax + 32 14 32 27 95
Send a message to Stephanie Delalieux
Birgen Haest
Tel. + 32 14 33 67 11
Fax + 32 14 32 27 95
Send a message to Birgen Haest
Back
