Determination of urban areas using region growth and fuzzy theory

• Grant number: 06/05839-5
• Support Opportunities: Scholarships in Brazil - Scientific Initiation
• Start date: April 01, 2007
• End date: March 31, 2008
• Field of knowledge: Physical Sciences and Mathematics - Computer Science
• Principal Investigator: Antonio Cesar Germano Martins
• Grantee: Diego Mendonça Arantes
• Host Institution: Universidade Estadual Paulista (UNESP). Campus Experimental de Sorocaba. Sorocaba , SP, Brazil

Abstract

Remote sensing aims at the study of Earth’s surface by the analysis of images acquired with equipments in airplanes, balloons or satellites. Environmental studies can be accomplished with remote sensing such as the evolution evaluation of urban area growth and the impact caused by natural phenomena or human intervention. For that, the implementation of image processing tasks in an automatic manner, such as segmentation and classification of urban areas or vegetation cover, is important. The more automatic those processes are, the more time will be reserved to the environmental specialist for the analysis of the processes taking place, as well as the application of remediation methods. In satellite images with 30 by 30 meters resolution, each pixel represents the sum of soil, vegetation and building characteristics. This way, pixel classification as belonging to one of those classes can be difficulty. For that, methods based on fuzzy logic can be useful. This research project proposes the study, implementation and test of digital process techniques that could be applied to satellite images for the segmentation of connected urban area. This can be done in a two step way: first the urban area is pre-segmented with classic region growing, and then a logic fuzzy algorithm can be used for the refinement of the border. To facilitate the application of this approach by environmental specialist, a friendly interface will be developed with C++ programing language. (AU)