Name
Abstract | ||
|---|---|---|
The Evaluation and Validation of CHRIS for National Forests Project (EVC) was selected by ESA's science team for their hyperspectral sensor,CHRIS as part of the PROBA mission. On September 2nd to 4th, 2006, a triplet acquisition over the Greater Victoria Watershed District (GVWD), taken in Mode-1, has provided us with 15 look angles in three consecutive days. The Minimum Zenith Angles (MZA) for each date were +20deg, -2deg and -23deg respectively, each of which has five nominal Fly-by Zenith Angles (FZA) of plusmn55deg, plusmn36deg and 0deg. This triplet has been processed and analyzed in order to assess the utility of CHRIS data for mapping forest parameters. Pre-processing is important in order to produce hyperspectral and directional data that is consistent spatially, spectrally, and directionally. Both the inherent horizontal and vertical noise have been reduced using the improved processes proposed by DIELMO 3D [1]. The GVWD has significant topographic relief that required a geometric correction based on a parametric approach [2], as implemented in PCI's Orthoengine. ATCOR 3 was used for the atmospheric correction. Further calibration was required, in which the atmospherically corrected data were matched to spectral ground data, measured during the overpass at angles matching the CHRIS images. This field calibration required the development of a simple goniometer and improved understanding of the true geometry of CHRIS image acquisitions, relative to observations made on the ground. A goal was to derive CHRIS algorithms for producing accurate estimates of forest parameters such as conifer forest species and biomass. The spectral information content provided information on the biochemistry of the forest canopy, while the multi-angle imagery gave information on the structural components of the forest canopy [3]. It has been shown that using multi-angle imagery can improve classification results of forest species by 8%, compared to the 80.3- % classification accuracy with only the nadir angles for CHRIS mode 4 data [4]. The triplet dataset offers improvements to our study of BRDF over this earlier study. The image triplet was acquired in mode 1, which provides more spectral bands; the images were acquired in a short time interval with very similar sun angles. |
Year | DOI | Venue |
|---|---|---|
2007 | 10.1109/IGARSS.2007.4423104 | Barcelona |
Keywords | Field | DocType |
image processing,remote sensing,topography (Earth),vegetation,AD 2006 09 02 to 04,ATCOR 3,CHRIS algorithms,CHRIS image acquisition,Compact High Resolution Imaging Spectrometer,DIELMO 3D,ESA science team,Fly-by Zenith Angles,Greater Victoria Watershed District,Minimum Zenith Angles,National Forests Project,PCI Orthoengine,Project for On Board Autonomy mission,atmospheric correction data,biochemistry,biomass,conifer forest species,forest attributes,forest canopy,forest parameter mapping,forest species,goniometer,hyperspectral sensor,multiangle imagery,spectral ground data,structural components,topographic relief,CHRIS,PROBA,forestry,hyperspectral,multi-angle,multi-date | Bidirectional reflectance distribution function,Nadir,Atmospheric correction,Tree canopy,Computer science,Remote sensing,Image processing,Hyperspectral imaging,Spectral bands,Zenith | Conference |
ISSN | ISBN | Citations |
2153-6996 | 978-1-4244-1212-9 | 1 |
PageRank | References | Authors |
0.40 | 5 | 5 |
Name | Order | Citations | PageRank |
|---|---|---|---|
| a dyk | 1 | 85 | 18.72 |
| David G. Goodenough | 2 | 84 | 19.70 |
| K. Olaf Niemann | 3 | 23 | 7.09 |
| Geordie Hobart | 4 | 31 | 6.44 |
| Hao Chen | 5 | 87 | 16.52 |