| dc.description.abstract | Floods are regular disasters that cause for the
greatest economic losses in the world. Due to that, victims
from flood disaster monitoring and the damage estimates
are important to the population, government authorities
also the insurance companies. Global demand for unman
aerial vehicles has increased significantly by the emerging
economics and it is capable enough to act it a vast area of
tasks which are considered dirty, dull, or dangerous by the
humans. UAVs provide a platform to minimize human
involvement and speediness the process of identifying and
locating the causalities, the mission of analyzing of
gathered images can be given to a computerized algorithm
which examines the images. Acomputerized algorithm
which can determine and geologically identify the
locations where flood victims can be originate is required
to identify the victims in a shorter time to direct rescue
teams towards them, it will reduce the death toll due to
flood disaster. This document describes the unique
solution built on hybrid networks and image processing by
using the aerial photographs. As a first innovation,
capturing the flood situation, converting and overlapping
image processing is used. Color and spatial information
extracted from simultaneous color matching and fractal
quality dimension matrixes are also used. The
experimental results of the real task finally show the
effectiveness of the method and the performance of the
algorithm. | |
