Characterization and surface reconstruction of objects in tomographic images of composite materials

Dissertação para obtenção do Grau de Mestre em Engenharia Informática

In the scope of the project Tomo-GPU supported by FCT / MCTES the aim is to build an interactive graphical environment that allows a Materials specialist to define their own programs for analysis of 3D tomographic images. This project aims to build a tool to characterize and investigate the identified objects, where the user can define search criteria such as size, orientation, bounding boxes, among others. All this processing will be done on a desktop computer equipped with a graphics card with some processing power. On the proposed solution the modules for characterizing objects, received from the identification phase, will be implemented using some existing software libraries, most notably the CGAL library. The characterization modules with bigger execution times will be implemented using OpenCL and GPUs. With this work the characterization and reconstruction of objects and their research can now be done on conventional machines, using GPUs to accelerate the most time-consuming computations. After the conclusion of this thesis, new tools that will help to improve the current development cycle of new materials will be available for Materials Science specialists.