Politecnico di Torino (logo)

Development of a real-time solution for an interactive VR representation of large star catalogues

Paolo Gallo

Development of a real-time solution for an interactive VR representation of large star catalogues.

Rel. Andrea Sanna, Eugenio Topa. Politecnico di Torino, Corso di laurea magistrale in Ingegneria Informatica (Computer Engineering), 2021

PDF (Tesi_di_laurea) - Tesi
Licenza: Creative Commons Attribution Non-commercial No Derivatives.

Download (7MB) | Preview
[img] Archive (ZIP) (Documenti_allegati) - Other
Licenza: Creative Commons Attribution Non-commercial No Derivatives.

Download (323kB)

This thesis takes place in the context of Virtual Reality and Data Visualization techniques applied to large astronomical datasets. The goal of this work is to improve and extend the already existing Astra Data Navigator application, built with the Unity game engine, and make it capable of loading and displaying large star catalogues in a realistic and real-time 3D environment. The software was built in the VR laboratory of ALTEC - Aerospace Logistics Technology Engineering Company - while interfacing with other European projects such as NEANIAS and ESA’s Gaia mission, which is particularly relevant to this work. The catalogue of celestial objects observed by the Gaia astrometric satellite is the largest collection of stars available to date, counting over 1.8 billion entries; being able to navigate and interact with this data in 3D would be extremely useful for both scientific and educational purposes, but most VR tools are limited to a much smaller object count and cannot be extended further. On the data management side, the application (which has an integrated star catalogue but also supports external data sources in the form of CSV files or SQL databases) offers the choice between two different modes: the user can choose to load all of the available data at startup and store it in the system memory, which requires more resources and increased loading time but then provides a seamless navigation of the 3D environment, or he can opt for a dynamic loading solution (better suited for large catalogues), that only selects relevant data based on the current observer position, saving a lot of resources but introducing additional loading times that interrupt the navigation experience. Processing and rendering such large amounts of objects in real-time can be a difficult task, which was however solved by carefully managing the application’s resources, choosing the right data structures, optimizing algorithms in order to take advantage of the multiple cores in modern processors and exploiting the massive parallelism provided by GPUs via programmable shaders. All the techniques employed to overcome these challenges are described in depth in this work, along with their implementation details and limitations, while also highlighting some of the features that are made possible by these systems. The results of performance analyses and user testing of the final product are also presented in this study, showing how most of the project’s goals have been successfully accomplished but also highlighting a few shortcomings of the implemented solution, which could be overcome by further development.

Relators: Andrea Sanna, Eugenio Topa
Academic year: 2020/21
Publication type: Electronic
Number of Pages: 82
Corso di laurea: Corso di laurea magistrale in Ingegneria Informatica (Computer Engineering)
Classe di laurea: New organization > Master science > LM-32 - COMPUTER SYSTEMS ENGINEERING
Aziende collaboratrici: Altec Spa
URI: http://webthesis.biblio.polito.it/id/eprint/18136
Modify record (reserved for operators) Modify record (reserved for operators)