Rokos, Demetrius
2 | Extension of
Linear (private roads, fences, etc.)
Spatial units of 2,000
Very high resolution remote
and forest areas
city plans and surface / spatial patterns | to 200 sqm size have | sensing images e.g. SPOT
(buildings, building plots, etc.) have | to meet the spectral | panchromatic images, KVR,
to be identified. requirements of the | IRS, IKONOS, and/or air-
spatial patterns. photos.
3 | Fencing of | Linear patterns have to be | None. KVR, IKONOS images,
private forests | identified. air-photos.
4 | Illegal building
Point or surface/spatial patterns
(depending on the size of the
building and image resolution) have
to be identified on a time series of
images to detect buildings' location
and reveal the time of their
construction.
Spatial units of 200
sqm size have to meet
the spectral
requirements of the
spatial patterns.
Very high resolution remote
sensing images (KVR,
IKONOS) and/or air-photos
taken in various years.
5 | Fragmentation
Linear (fences) patterns and/or
spatial/surface ^ patterns with
spectral characteristics different
from those of the surrounding area
have to be identified.
Spatial ^ units of
500sqm size at least.
KVR, IKONOS images, air-
photos for linear patterns.
SPOT XS, IRS for spatial
patterns.
6 | Quarries
Spatial/surface patterns have to be
recognized on a time series of
remote sensing images to detect
Various as required,
depending on quarry
size.
LANDSAT TM, SPOT XS,
IRS images, air-photos taken
in various years.
quarry location and activities.
7 | Reforestation of | Surface/spatial patterns ~~ with | Spatial units of | SPOT XS, LANDSAT TM,
burnt forest specific spectral signature have to | 1,000sqm size at | IRS, IKONOS multispectral
areas be recognized. least. images and air-photos taken
in various years.
8 | Building Surface / spatial patterns have to be | Spatial units of Air-photos taken before
Associations identified. 200sqm size at least | 1974 and very high
have to meet the | resolution remote sensing
spectral requirements | images (KVR1000,
of the spatial | IKONOS) and/or air-photos
patterns. taken after this year.
Table 3. Objects/patterns to be detected. Spatial, spectral, temporal requirements for EO data.
4.3 Remote Sensing Methods and Techniques
4.3.1 Pre-processing methods and techniques. Among pr
considered mandatory. Radiometric correction, topographic norma
e-processing methods, the geometric correction is
lization and atmospheric correction could be
considered optional, depending on the image quality and on topographic and atmospheric conditions.
4.3.2 Image processing methods and techniques. These are listed in the table 4 (Rokos, 1988, 1994b).
Category Image processing methods and techniques
1 Forests and
forest areas
characterization
1945.
photos.
vegetation.
natural vegetation.
e Application of photo-interpretation methods and techniques on the air-photos taken in
° Application of photo-interpretation methods and techniques on the most recent air-
e Application of vegetation indices on multispectral satellite imagery to identify natural
e Application of classification algorithms on multispectral satellite imagery to identify
e Application of change detection algorithms on two sets of remote sensing data to
indicate temporal changes in forests and forest areas.
1272
International Archives of Photogrammetry and Remote Sensing. Vol. XXXIII, Part B7. Amsterdam 2000.