Modelagem 3d de áreas urbanas utilizando processamento digital de imagem

Abstract

On one hand it is increasing the attention to urban planning, especially in urban expansion, planning and taxes. Several studies were performed with the purpose of transforming urban planning in a more scientific way to improve urban planning. Thus, the governments are demanding the use of more intelligent mechanisms in the management of the cities. In Brazil this is no different. Through the City Statute Law, the Brazilian government is demanding that the cities present some urban planning mechanisms. Although this law and the guides, offered by the Brazilian government, indicate the mechanisms to develop urban planning they don’t described the tools or the methods to develop the analysis that is demanded. On the other hand, the vision is the main mechanism that humans use to recognize the world around them in a way that it is considered one of the most important sense that people have. Is through this sense that one can recognize, classify and differentiate the things that are around. As the vision being a so important sense, it is interesting give this ability to computers and machines, making them able to identify the environment as well as human beings identify. By the union of the need for new tools to contribute to the urban planning and the importance that vision has in carrying out important actions emerges the objective of this work: to provide an additional tool using computer vision technologies to the urban planning, providing more information for directors of cities, urban designers and urban planners about the volumes of urban areas. In other words, this work presents how to take advantage of 2D terrestrial photographs of urban areas to get 3D models of urban structures. Thus, contributing as an additional tool for the development better urban master plans and regional plans contributing to a more sustainable growth of cities.