Development of an Algorithm to Identify the Locations of Flood Victims Using Digital Image Processing

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.