high resolution (H-res), 147
homogeneous, 147
low resolution (L-res), 148
Image resampling, 5
Image segmentation, 5, 7, 171, 175–176,
187–188, 197–199, 201, 202, 205,
268, 306
algorithms, 191, 197–199
edge-based methods, 198
histogram thresholding, 197
image feature space clustering, 197
region-based approaches, 198
based on mean-shift segmentation and
FNEA, 202
development of estimation of scale
parameter (ESP), 203
edge-based, 171, 175, 197
multiscale, 202, 206
scale parameters, based on evaluation
approach, 204
Image texture analysis, 146
Indianapolis, 233
efficiency of fractal measurement, for
detecting scaling properties, 250
fractal analysis
for temporal change
characterization, 248–249
using LULC classes, 242–243
using LULC maps, 239
using raw images, 237–239
using raw red bands classified by LULC
classes, 239–242
using resampled raw red images,
242–248
IKONOS data, 234
information of satellite images, 235–236
Lacunarity analysis, use of, 250
landscape characterization, at multiple
scales by fractal
measurement, 236–237
pixel aggregation levels, 250
sensor’s nominal spatial resolution,
249–250
total area, 233
total population, 233
use of fragmented raw red bands filtered
by LULC types, 249
using traditional ISODATA unsupervised
algorithm, 250
values of root meansquare error
(RMSE), 234
Individual-population-communityecosystem, 146
Isotropic multiresolution analysis, 173
Kernel-based classification, 181–182
K-nearest neighbors, to map forest
carbon, 108
Koffler Scientific Reserve
chlorophyll data, and interpretations
canopy-level chlorophyll content,
132
CHRIS/PROBA data, 135
data-derived spectral index vs. content
at canopy level, 132–133
field hyperspectral data–derived SR vs.
content at landscape level,
133–134
lab-based spectral reflectance, 131
relating remote sensing data to, 131
common species found in, 128
field data collection, 128–130
QuickBird image acquisition and
preprocessing, 130
less regression samples, with large
variation in, 135
variation in species percentage, due to
scaling data, 135
scaling up from leaf to canopy, and
landscape levels, 130–131
spectral indices, 130
study sites, 128
LAI. See Leaf area index (LAI)
Land–atmosphere energy
exchange processes, 37
fluxes, 37
Land cover, defined, 62
Land cover extraction
hyperspectral remote sensing, 302
multi- and hyperspectral sensors,
spectral properties, 303–304
overview, 302, 305
from remote sensing, importance of
observation scale in, 310–311
Land cover mapping, 8, 62–63, 291
accuracy assessment, 307–308
kappa coefficients, 311
INDEX
327
homogeneous, 147
low resolution (L-res), 148
Image resampling, 5
Image segmentation, 5, 7, 171, 175–176,
187–188, 197–199, 201, 202, 205,
268, 306
algorithms, 191, 197–199
edge-based methods, 198
histogram thresholding, 197
image feature space clustering, 197
region-based approaches, 198
based on mean-shift segmentation and
FNEA, 202
development of estimation of scale
parameter (ESP), 203
edge-based, 171, 175, 197
multiscale, 202, 206
scale parameters, based on evaluation
approach, 204
Image texture analysis, 146
Indianapolis, 233
efficiency of fractal measurement, for
detecting scaling properties, 250
fractal analysis
for temporal change
characterization, 248–249
using LULC classes, 242–243
using LULC maps, 239
using raw images, 237–239
using raw red bands classified by LULC
classes, 239–242
using resampled raw red images,
242–248
IKONOS data, 234
information of satellite images, 235–236
Lacunarity analysis, use of, 250
landscape characterization, at multiple
scales by fractal
measurement, 236–237
pixel aggregation levels, 250
sensor’s nominal spatial resolution,
249–250
total area, 233
total population, 233
use of fragmented raw red bands filtered
by LULC types, 249
using traditional ISODATA unsupervised
algorithm, 250
values of root meansquare error
(RMSE), 234
Individual-population-communityecosystem, 146
Isotropic multiresolution analysis, 173
Kernel-based classification, 181–182
K-nearest neighbors, to map forest
carbon, 108
Koffler Scientific Reserve
chlorophyll data, and interpretations
canopy-level chlorophyll content,
132
CHRIS/PROBA data, 135
data-derived spectral index vs. content
at canopy level, 132–133
field hyperspectral data–derived SR vs.
content at landscape level,
133–134
lab-based spectral reflectance, 131
relating remote sensing data to, 131
common species found in, 128
field data collection, 128–130
QuickBird image acquisition and
preprocessing, 130
less regression samples, with large
variation in, 135
variation in species percentage, due to
scaling data, 135
scaling up from leaf to canopy, and
landscape levels, 130–131
spectral indices, 130
study sites, 128
LAI. See Leaf area index (LAI)
Land–atmosphere energy
exchange processes, 37
fluxes, 37
Land cover, defined, 62
Land cover extraction
hyperspectral remote sensing, 302
multi- and hyperspectral sensors,
spectral properties, 303–304
overview, 302, 305
from remote sensing, importance of
observation scale in, 310–311
Land cover mapping, 8, 62–63, 291
accuracy assessment, 307–308
kappa coefficients, 311
INDEX
327