Interfacement of the methodology developed to reconstruct features partially detected in digital images to cartomorph and development of methodology for minimizing noise around the rebuilded features

• Grant number: 18/21965-8
• Support Opportunities: Scholarships abroad - Research Internship - Scientific Initiation
• Start date: December 08, 2018
• End date: March 07, 2019
• Field of knowledge: Physical Sciences and Mathematics - Geosciences
• Principal Investigator: Erivaldo Antonio da Silva
• Grantee: Ana Luisa Chaves Figueira
• Supervisor: Pedro Miguel Berardo Duarte Pina
• Host Institution: Faculdade de Ciências e Tecnologia (FCT). Universidade Estadual Paulista (UNESP). Campus de Presidente Prudente. Presidente Prudente , SP, Brazil
• Institution abroad: Instituto Superior Técnico (IST), Portugal
• Associated to the scholarship: 17/13029-8 - Development of methodology for features partially detected reconstruction after cartographic features extraction in digital images: case study: curved features, BP.IC

Abstract

Extraction processes of cartographic interest features from digital images have been of great importance in the area of Cartography. These processes have as main focus the identification of targets present in the terrestrial surface, as, for example, highways, drainage, runways of airports, and the updating of cartographic products. The land surface is under constant changes caused mainly by anthropic actions and nature, and the identification of such changes is essential for the cartography of a country with the continental dimensions of Brazil. Several techniques of Digital Image Processing can be used in the process of extraction of features, and, in most cases, the results obtained by these techniques have features partially detected, causing a loss in the quality of the extraction process. There are studies that involve the reconstruction of such features, but their applications in curvilinear features produce a quality effect that is lower than expected. Thus, based on what has elapsed, the present project has as main objective to develop a methodology for the reconstruction of curvilinear features partially detected through the use of techniques based on inpainting, aiming the improvement of the results in the process of extraction of cartographic features in images digital images. The tasks of interfacing the selected techniques, implemented and tested with the CARTOMORP library, as well as proposing a methodology to minimize the noises present in the surroundings of the features of interest will also be performed. (AU)