A framework for the management of deformable moving objects

• Autor(a) principal: Duarte, José Horácio Fradique
• Data de Publicação: 2017
• Tipo de documento: Dissertação
• Idioma: eng
• Título da fonte: Repositório Científico de Acesso Aberto de Portugal (Repositórios Cientìficos)
• Texto Completo: http://hdl.handle.net/10773/24805
• topic: Moving Objects; Spatiotemporal Data Models; Spatiotemporal Databases; Compatible Triangulation; Rigid Interpolation

There is an emergence of a growing number of applications and services based on spatiotemporal data in the most diverse areas of knowledge and human activity. The Internet of Things (IoT), the emergence of technologies that make it possible to collect information about the evolution of real world phenomena and the widespread use of devices that can use the Global Positioning System (GPS), such as smartphones and navigation systems, suggest that the volume and value of these data will increase significantly in the future. It is necessary to develop tools capable of extracting knowledge from these data and for this it is necessary to manage them: represent, manipulate, analyze and store, in an efficient way. But this data can be complex, its management is not trivial and there is not yet a complete system capable of performing this task. Works on moving points, that represent the position of objects over time, are frequent in the literature. On the contrary there are much less solutions for the representation of moving regions, that represent the continuous changes in position, shape and extent of objects over time, e.g., storms, fires and icebergs. The representation of the evolution of moving regions is complex and requires the use of more elaborate techniques, e.g., morphing and interpolation techniques, capable of producing realistic and geometrically valid representations. In this dissertation we present and propose a data model for moving objects (moving points and moving regions), in particular for moving regions, based on the concept of mesh and compatible triangulation and rigid interpolation methods. This model was implemented in a framework that is not client or application dependent and we also implemented a spatiotemporal extension for PostgreSQL that uses this framework to manipulate and analyze moving objects, as a proof of concept that our framework works with real applications. The tests’ results using real data, obtained from satellite images of the evolution of 2 icebergs over time, show that our data model works. Besides the results obtained one important contribution of this work is the development of a basic framework for moving objects that can be used as a basis for further investigation in this area. A few problems still remain that must be further studied and analyzed, in particular, the ones that were found when using the compatible triangulation and rigid interpolation methods with real data.