文档介绍：ANALYSIS AND CLASSIFICATION OF REMOTELY
SENSED SATELLITE DATA USING WORKS
AND FUZZY LOGIC SYSTEMS TECHNIQUES
Sharath Tadepalli
Graduate student, Geomatics area, School of Civil Engineering, Purdue University
Student Member, ACSM, Email: ******@
ABSTRACT: acquired by a device that is not in contact
with the object, area or phenomenon[2].
This paper aims at using the various The data collected may be spatial, spectral
algorithm based work or temporal in nature and the features thus
techniques and reasoning and logic established may be represented in the
based fuzzy systems theory to analyze geo image space, feature space or the signal
spatial remote sensing satellite data for space. However in this study we
the purposes of data interpretation, concentrate preferably on the feature space
pattern recognition, error analysis and representation.
classification. A part of the project is thus
dedicated to the understanding and work is defined as a learning
implementation of these techniques and process controlled by a learning algorithm,
analyze their relevance and suitability to the function of which is to modify the
the above mentioned exercises. A brief synaptic weights of work in an
comparative study is then carried out to orderly fashion to attain a desired design
benchmark the results parison objective. Among the various benefits of
with various conventional tools used in works we concentrate on the
this regard. The usefulness of statistical input-output mapping aspect for the
pattern recognition methods including implementation of the project objective.
feature selection, feature extraction,
decision boundary analysis and data Other benefits of works that are
classification are also looked at in both extensively used implicitly in this study
the classical sense as well as from a include non-linearity, adaptivity,
trainable work point of view. evidential response, contextual
In concl