Digital image processing techniques for the identification of the appropriate flood areas for the malaria vector procreation

Abstract

The present work intends to identify the areas which have appropriate environment for the procreation of the disease vector in the northern part of Labrea municipality, in the state of Amazonas. The environmental conditions found in the study region are highly favorable to the elevate density of mosquitoes and to their procreation in the larvae phase, as it is known that the ideal environment considered to its procreation is the aquatic one, where occurs around 3⁄4 of the mosquito life cycle. The digital image processing techniques have been considered as an important tool to identify these areas and have aided the decision making in the public health segment. In order to achieve the purpose of this work, two approaches were made: the object-oriented analysis classification of an optical image and the proposal of a new methodological approach for the terrain description, resulting from the application of the algorithm called HAND. In this approach, the purpose is to characterize both the water bodies identified during the image classification and the possible flood regions, thus, favorable for the procreation of the Anopheles vector. In the spatial analysis stage, a local spatial regression model was used. As a result, it was concluded that de descriptor terrain model in the flood areas of the study region is an explanatory factor of the malaria incidence in the northern part of Labrea.