236 
THE METHOD OF GARBAGE REPOSITORY DETECTION AND CONTROL 
Ivana Javorovi6, Toni Kova6, Boris Smiljanic 
Remote Sensing Laboratory 
llica 242, Zagreb 
Croatia 
Commission VII, Working group 3 
KEY WORDS: garbage repository, unsupervised classification, GIS 
ABSTRACT 
The process of unsupervised classification in order to separate essential detection characteristics of 
garbage repository near urban environments has been applied. The classes were defined upon known 
garbage repository locations. Multisensor merging data of Landsat TM and SPOT panchromatic 
images are used for the classification and for more precise spatial locating in database. The method 
can as well be applied to detect wide uncontrolled repository in order to supervise the environmental 
polluting. A GIS system has been developed for storing the classified results with the possibility of 
monitoring modifications in the course of time. 
RÉSUMÉ 
Le proces de la classification non supervisée est appliqué pour separer les caractéristiques 
essentielles pour la découverte du dépot d'ordures à cote de l'environnement urbain. Les classes sont 
definies sur les dépot d'ordures connus. La fusion multisensorique des données des images Landsat 
TM et SPOT pancromatique, sont employés pour la classification et pour l'emplacement spatial plus 
précis dans la base de données. Cette méthode peut étre employée pour découvrir des grands dépot 
d'ordures non cóntrole et pour surveiller la pollution de l'environnement. Le SIG système est développé 
pour noter des résultats qui classifiés avec la possibilité de la modification au cours de temps. 
INTRODUCTION 
Growth of the human population and rapid 
technological and economical developments of 
the human society inevitably increase danger 
for the nature. Unadequate garbage 
managing, particularly garbage repositories, 
could have enormous influence on the 
ecology. Remote sensing methods and GIS 
technology is proposed as a tool for the 
locating of the “wild” garbage repositories, and 
for the controlling of the growth in the course 
of time for known and wild ones. 
For the interpretation and classification 
methods SPOT panchromatic and Landsat TM 
images of the area containing known garbage 
repositories near urban parts were used. 
ERDAS IMAGINE software version 8.3 is used - 
for the interpretation. GIS system is 
developed on ARC/INFO 6.1.1. software 
platform. 
METHOD 
Garbage repositories with large dimensions 
are specific for the intensive organic 
disintegration, which results with emission of 
the heat. This is a starting point for the 
classification. Such areas can be realtively 
easy recognized on the satellite images in the 
infrared and thermal bandwidths. We merged 
the Landsat TM and SPOT panchromatic 
image data, which have better resolution. This 
data combination takes advantage of the 
spectral resolution of the Landsat TM data and 
the spatial resolution of the SPOT data. Also, 
this data combination is a basis for the 
classification and for more precise spatial 
locating in database. 
We used different type of data for 
classification. The first type of data contains 
hydrothermal composite (the indices process) 
and TM band 6. The indices process used to 
create output images by mathematically 
combining the digital number values of 
different bands. The output hydrothermal 
composite contains three bands with the 
following output values: 
1. bandS/band7 
2. band3/band1 
3. band4/band3. 
The four bands source of data of certain area 
is defined on known garbage repository 
International Archives of Photogrammetry and Remote Sensing. Vol. XXXII, Part 7, Budapest, 1998