60
Blue Green Red
T
50
40 r1
10
Percent of Reflectance
Dead Vegetat
30 + geta ien ee CS
Reflected-Infrared
Der.
—
Healthy Vegetation
ae 2er
0.4 0.5 0.6: 0.7 0.8 dH 1.1
Wavelength (Micrometers)
Fig.3 Typical Characteristics of Spectral Reflectance
of TM band 3, 4 brightness values are used
for this training. Calculated index is
called Normarized Difference (Vegetation)
Index (i.e. NVI or NDi).
(e) Density Slicing
The density slicing is one of the
basic techniques for an image enhancement.
Density slicing is the conversion of the
continuous tone of digital image data to
some classes. Every interval of each class
corresponds to the specific brightness
value range. Used personal computer system
has basically only 8 colors for displaying
on CRT monitor. By mentioned density
slicing and histogram measurement for NVI
data, basic concept of NVI and contrast
stretch can be understood.
3.2 Basic Principles of
Geographic Information System
Geographic information system (GIS)
is a term that is to store the geographi-
cally distributed informations in XY
coordinate system, retrieval and analysis
of these informations using computer
system. It also contains database, the
hardware and software, retrieving, analyz-
ing, and displaying.
(a) Data Structure
The geographic location data that are
acquired by a digitizer is classified to
following three kinds of data feature
types; (1)Point, (2)Line, (3)Polygon.
Every location data also has an
attribute information for its explanation
such as one's name. The point type data is
a single point location data and expressed
by xy coordinate. The line type data is
the link of points data and its data
consists of same structure as point type
data. The polygon type data is almost same
as the line type data. But this kind of
data constitutes one closed loop. Data
structure is the same as line type data.
Fig.4 shows the point, line and polygon
type data structures. It is important to
manage the point feature because point
data is the basic for all kinds of data
types.
(b) Vector and Raster Data Type
Generally, the data type of GIS data
is classified to two kinds of data model.
One is the vector data model and other is
the raster data model.
Typical raster data is a digital
image data such as remote sensing data. In
the raster data model, one homogeneous
unit is expressed by cells (pixels) that
consists of grid-cells on the ground.
Actual grid-cell creates a matrix that is
superimposed on the earth surface. And
within this systematic array of grid-cell,
the attribute data is collected in order
to produce raster type data.
The vector data model provides a more
compact data structure than the raster
model. But data structure is more complex
than the raster type. Both manipulation
and enhancement of digital images can not
be effective for the vector domain.
er
Log
Location
> Point XY
» Line XY Xs c» XSYà
ly Polygon X1,Y1,X9,Yg, Xg, Yg, X4, Y4
Fig.4 GIS Data Structure
(c) Geometric Manipulation
The original satellite image coordi-
nate can not be overlapped on XY coordi-
nate system of the map. However remote
sensing data can be projected on some map
coordinate using the relationship between
image data coordinate and map coordinate.
GIS data has a capability to change
scale, change map projection, remove,
distortion and execute coordinate rotation
and translation depend on its necessity.
Fig.5 shows the processes are called
fundamentals of geometric manipulation.
4. Application Training
In this training, each practice is
basic techniques for remote sensing and
geographic information system. However all
practices could be constructed one train-
ing course.
The analysis area of this application
training is Hachioji. In this area, re-