inbul 2004
:rungsbüro
1 F., 2003.
ges Using
ley Piater
r Science,
162-171.
lung und
‚getragenen
(PFIFF),
ommission,
nal forum;
ion digital
| Resources
try Centre,
38. Snakes:
f Computer
r-Data with
ertation der
sitüt Bonn,
us digitalen
Geodátische
^ Analysis:
Heidelberg
ar structures,
1d Machine
Bäumen aus
Deutsche
002. GIS in
atics, No. 5,
ld Extraction
1 Algorithm,
note Sensing,
MULTISPECTRAL ANALYSIS OF SATELLITE IMAGES
Nour el islam. BACHARI Salim .KHODJA, and Ahmed.Hafid. BELBACHIR
Laboratoire d'Analvse et d'Application des Rayonnements (LAAR)
Département de Physique, Université de Sciences et de la Technologie d'Oran (USTO)
B.P. 1505 EI-M'Nouar, Oran, 31000 Algeria. Bacharil 0@yahoo. fr
ABSTRACT :
Spectroscopy by satellite images has many new and exciting applications. Areas such as
many others have great potential for new research advances using this technology. There
that need to be addressed on a regular basis in order to achieve valuable research results.
This work presents a multispectral analysis of natural targets behavior of a study area, using the TM sensor images (Thematic
Mapper) on board the satellite LANDSAT 5. The spectral classification was done according to the following steps :
a) spectral reduction by the Minimum Noise Fraction (MNF) transformation,
b) spatial reduction by the Pixel Purity Index (PPI) and
agriculture, forestry, geology and
are several common issues though,
C) manual identification of the endmembers using the N-dimensional visualizer.
These procedures allowed showing the kind of landscape that presents a strong correlation with geology.
Keywords : Landast TM Multispectral Images, MNF transformation, PPI, endmembers, Reflectance spectrum, ENVI
software
1- INTRODUCTION by its 6 channels images TM (Thematic Mapper 1, 2, 3,
4, 5, 7). The infrared thermal channel TM6 will not be
treated in our study. Here we will create Color-Ratio-
Composite (CRC) Images using standard TM band-ratio
images. This method tries to get around the limitations of
relatively broad spectral bands in Landsat TM data by
using ratios of bands to determine relative spectral slope
between bands and thus the approximate shape of the
spectral signature for each pixel. Common band-ratios
With a multispectral analysis of an images collected by
radiometers and detectors embarked on a satellite, one
can make : " spectroscopy by satellite imagery ".
Imaging spectrometers or "multispectral sensors" are
remote sensing instruments that combine the spatial
presentation of an imaging sensor with the analytical
capabilities of a spectrometer. With a radiometer such as
AVIRIS ^" Airborne Visible/Infrared Imaging include: Band-Ratio 5/7 for Clays, Carbonates,
Vegetation; Band-Ratio 3/1 for Iron Oxide; Band-Ratios
2/4 or 3/4 for Vegetation; and Band-Ratio 5/4 also for
vegetation.
Spectrometer " where the number of spectral bands
exceeds the 200, and with a spectral resolution on the
order of 10 nm or narrow, it can produce a complete
spectrum for each pixel of the image. the case of this
radiometer enters on the "Hyperspectral" field. In the
case of our study, we are using the radiometer TM
(Thematic Mapper), which is on board the satellite
Landsat 5, and which only has 7 multispectral bands,
With a spectral resolution on the order of 100 nm. Like
finale result, one using the spectroscopy by satellite
imagery, we can identify the spectra of various materials
which are on various terrestrial surfaces.
However there are several common issues though, that
need to be addressed on a regular basis in order to
achieve valuable research results. One always needs high
quality data, the proper radiometric, geometric and
atmospheric corrections, and appropriate analysis
techniques in order to achieve acceptable results. The
purpose of this work is to present and explain the main
analysis techniques available for multispectral imagery.
This will be done using ENVI 3.1 software (the
Environment for Visualizing Images), which is a well-
known hyperspectral analysis program. The analysis
techniques will include :
* minimum noise transformation,
* pixel purity index and
* n-dimensional visualization.
2.DATA TO ANALYZE
Figure.1. shows the combination of 5/7, 3/1, 2/4 (RGB)
results in an image in which clays/carbonates are
magenta, iron oxides are green, and vegetation is red.
Other ratio combinations and color schemes can be
designed to highlight specific materials.
3. ANALYSIS
In our Study, we will treat a satellite image representing a
part of the region of Laghouat (Algeria), taken by the
Satellite Landstat 5 on April 12, 2001. It is represented
Since the goal of our work is to extract a reflectance
spectrum, the treated image must be calibrated in
1071
