54
J.Jensen
sification logic to extract thematic information from remotely sense data (Jensen
and Qiu, 1998). However, this is changing. The terrain usually grades from one
land cover into another without 'hard' partitions. In fact, the 'fuzzy transition
interface' between homogeneous terrain elements is often where the greatest species diversity of plants .and animals exists. Therefore, 'fuzzy' classification algorithms will likely see dramatic utility in the future.
In the past decade there has been a tremendous increase in the amount of active
microwave (radar) imagery available for hydrologic investigations. The launch of
the Canadian RADARS AT in 1995 stimulated research even more. Only a few
image processing systems provide the software necessary to remove or adjust the
speckle associated with the raw radar data and to geometrically warp it into a
ground-range map projection (as opposed to the original slant-range geometry).
3.2.4 Pbotogrammetric Information Extraction
This is one of the most exciting areas of digital image processing. For 50 years it
has been necessary to use expensive and time consuming analytical stereoplotters
to a) extract digital elevation models from stereoscopic vertical aerial photography, and b) produce orthophotographs. With advances in desktop computers and
soft-copy photogrammetry software, it is now possible to extractDEMs from both
aerial photography (Fig. 3.5) and satellite digital data and then produce accurate
orthoimages (e.g., Greve et al., 1992; Jensen, 1995b; 1997). Some of the new
software can produce true orthophotos where the building footprint is in its proper
planimetric location over the foundation and all relief displacement has been removed (Fig. 3.6).
Hydrologists, natural resource managers, and water management groups can
now produce accurate DEMs and orthophotos on demand for their local modeling
purposes rather than being dependent on the dreadfully slow cycle of government
DEM and orthophoto production. This should be of significant value for urban
hydrologic studies of pervious versus impervious areas and their elevation, slope,
and aspect. Appendix 3.1 identifies several vendors that provide photogrammetric
image processing software.
3.2.5 Metadata and ImageJMap Lineage Documentation
Metadata is data about data. Many countries have adopted rigorous national standards about the content, accuracy, and transmission associated with map and image spatial data. In the United States, the Federal Geographic Data Committee
(FGDC) has developed stringent metadata standards for all data produced for
government use. All digital image processing systems and GIS in the future will
eventually provide detailed metadata and image lineage (genealogy) information
about the processing applied to each image or map (Lunetta et al., 1991; Lanter,
and Veregin, 1992; Jensen and Narumalani, 1992; FGDC, 1998). The image lineage information is indispensable when the products derived from the analysis of
remotely sensed data are subjected to intense public scrutiny or litigation.
J.Jensen
sification logic to extract thematic information from remotely sense data (Jensen
and Qiu, 1998). However, this is changing. The terrain usually grades from one
land cover into another without 'hard' partitions. In fact, the 'fuzzy transition
interface' between homogeneous terrain elements is often where the greatest species diversity of plants .and animals exists. Therefore, 'fuzzy' classification algorithms will likely see dramatic utility in the future.
In the past decade there has been a tremendous increase in the amount of active
microwave (radar) imagery available for hydrologic investigations. The launch of
the Canadian RADARS AT in 1995 stimulated research even more. Only a few
image processing systems provide the software necessary to remove or adjust the
speckle associated with the raw radar data and to geometrically warp it into a
ground-range map projection (as opposed to the original slant-range geometry).
3.2.4 Pbotogrammetric Information Extraction
This is one of the most exciting areas of digital image processing. For 50 years it
has been necessary to use expensive and time consuming analytical stereoplotters
to a) extract digital elevation models from stereoscopic vertical aerial photography, and b) produce orthophotographs. With advances in desktop computers and
soft-copy photogrammetry software, it is now possible to extractDEMs from both
aerial photography (Fig. 3.5) and satellite digital data and then produce accurate
orthoimages (e.g., Greve et al., 1992; Jensen, 1995b; 1997). Some of the new
software can produce true orthophotos where the building footprint is in its proper
planimetric location over the foundation and all relief displacement has been removed (Fig. 3.6).
Hydrologists, natural resource managers, and water management groups can
now produce accurate DEMs and orthophotos on demand for their local modeling
purposes rather than being dependent on the dreadfully slow cycle of government
DEM and orthophoto production. This should be of significant value for urban
hydrologic studies of pervious versus impervious areas and their elevation, slope,
and aspect. Appendix 3.1 identifies several vendors that provide photogrammetric
image processing software.
3.2.5 Metadata and ImageJMap Lineage Documentation
Metadata is data about data. Many countries have adopted rigorous national standards about the content, accuracy, and transmission associated with map and image spatial data. In the United States, the Federal Geographic Data Committee
(FGDC) has developed stringent metadata standards for all data produced for
government use. All digital image processing systems and GIS in the future will
eventually provide detailed metadata and image lineage (genealogy) information
about the processing applied to each image or map (Lunetta et al., 1991; Lanter,
and Veregin, 1992; Jensen and Narumalani, 1992; FGDC, 1998). The image lineage information is indispensable when the products derived from the analysis of
remotely sensed data are subjected to intense public scrutiny or litigation.