Page 171 - Digital Analysis of Remotely Sensed Imagery

136 Cha pte r F o u r

offers limited capabilities of integrating remote sensing data with
non-remote sensing vector data. Spatial analysis functions that have
become an integral part of standard image analysis are also absent
from this package.

4.7 GRASS
The Geographic Resources Analysis Support System, commonly known
as GRASS, is a computer system for image processing, spatial modeling,
and visualization of many types of data. This free system was initially
developed by the U.S. Army Construction Engineering Research Labora-
tories as a tool for land management and environmental planning by the
military. It has evolved into a powerful system for a wide range of appli-
cations in many different areas, both academic and commercial (Shepard,
2000). Unlike all other image processing systems which are commercially
oriented, this open source system allows users to make their own contri-
butions to perfect it via a sophisticated library. Although a raster GIS
primarily, GRASS does have an image processing component apart from
geospatial data management and analysis, graphics/maps production,
spatial modeling, and visualization. There are more than 350 programs
and tools to render maps and images, process multispectral image data,
and manipulate and manage spatial data. Of these programs, 26 are
designed specially for image processing, such as data preparation (image
rectification and mosaicking of up to four images), image transformation
(principal component analysis, canonical components analysis, Tasseled
Cap (Kauth-Thomas) transformation, fast Fourier transformation, hue-
intensity-saturation to red-green-blue color transformation and vice
versa), image georeferencing, and image classification (clustering analy-
sis, maximum likelihood, contextual image classification).
GRASS modules are organized into display, general file manage-
ment, imagery (satellite data), photo, postscript (for map printing),
raster shell scripts (macro), sites (for point data), vector (for digitizing
and making aesthetically pleasing maps), and 3D visualization, such
as fly-through animations (Shepard, 2000). GRASS can be run on a
plethora of platforms and interface with databases to develop new
data and manage existing data. Existing data in a variety of formats
can be read by GRASS, such as ArcGIS and MapInfo. Besides, graphic
formats and even ASCII text can be imported and exported between
GRASS and other systems. In its latest release (version 6.3), more than
30 modules have been added to GRASS to broaden its functionality,
such as atmospheric correction and importing and exporting attribute
tables in various formats. In particular, its ability to handle vector
data has improved considerably, such as vector editing, generalization,
and importing/exporting.
Written in the C language and operational in the UNIX environ-
ment, GRASS used to be exclusively command driven. This mode
of interface has been replaced with the introduction of the GIS