Page 158 - Digital Analysis of Remotely Sensed Imagery

Image Pr ocessing Systems 125

Underneath the panel are many icons that provide a shortcut to access
some functions directly without going through the menu.
ER Mapper supports an extensive range of image formats in
which popular satellite data are saved, such as Landsat MSS, SPOT,
and AVHRR. Major proprietary image formats such as ERDAS Imag-
ine, vector data (i.e., ArcInfo coverage and MicroStation DGN), and
non-remote sensing DEM data can all be imported into ER Mapper or
exported to other systems (ER Mapper, 2007). Images can be directly
read and saved in several common formats, such as Joint Photogra-
phers Experts Group (JPEG), JPEG 2000, and GeoTIFF. Processed
results can be exported to these formats, as well. Furthermore, those
image formats that are directly importable can be imported through
generic import functions. In addition to remote sensing data, ER
Mapper also has tools for analyzing DEM data, such as conversion of
point and line data into a raster DEM from which topographic param-
eters (e.g., aspect, gradient, shaded relief, and 3D shading) may be
produced. Specialized functions are also available for processing
radar data, from which speckle noise may be removed and texture
derived. Output of processed results is via the map composition and
printing tools. All final results may be further embellished using the
map composition functions. Both simple and complex maps can be
composed using a library of predefined postscript map objects, such
as legends, coordinate grids, scale bars, north arrows, color bars, and
symbols.
As a major data preparation step, image georeferencing may be
based on one of the seven models, including polynomial, triangula-
tion, and map-to-map reprojection using GCPs derived from other
images or maps. Images can be georeferenced to a known ground
coordinate system using the geocoding tool. Large-scale aerial photo-
graphs can be orthorectified using GCPs collected from georeferenced
images or using the exterior orientation parameters of the camera.
Georeferenced images may be reprojected from one datum to another
or resampled to a different spatial resolution using the neareast neigh-
bor, bilinear, or cubic convolution method. Multiple georeferenced
images up to 100 may be stitched together to form a seamless mosaic
after their color or tone has been calibrated using the powerful wiz-
ards, or their contrast matched through histogram matching. The
mosaicked image may be subset with a vector polygon to any political
boundary or the boundary of the study area. Images from different
sensors may be fused to create a best view of the area under study.

4.4.2 Image Display
Image display is via ER Viewer, a free and easy-to-use tool that allows
interactive roaming and zooming of very large images. ER Mapper
has a powerful toolset for many kinds of display. One image may be
displayed with different options. For instance, a raster image may be
displayed as transparent, or over a shaded DEM. The impact of solar