Integrating Gaussian Splatting with Semantic Labels for Heritage BIM

Student Report (2024)
Authors

M.M. van Arnhem (TU Delft - Architecture and the Built Environment)

Q. YANG (TU Delft - Architecture and the Built Environment)

S.R.H.W. Tew (TU Delft - Architecture and the Built Environment)

X. Zhao (TU Delft - Architecture and the Built Environment)

W.H.J. Kahn (TU Delft - Architecture and the Built Environment)

Supervisors

E. Verbree (TU Delft - Digital Technologies)

Y.Y. Yu (TU Delft - Digital Technologies)

Florent Poux ()

Faculty
Architecture and the Built Environment
More Info
expand_more
Publication Year
2024
Language
English
Graduation Date
15-11-2024
Awarding Institution
Delft University of Technology
Programme
Geomatics
Faculty
Architecture and the Built Environment
Reuse Rights

Other than for strictly personal use, it is not permitted to download, forward or distribute the text or part of it, without the consent of the author(s) and/or copyright holder(s), unless the work is under an open content license such as Creative Commons.

Abstract

In recent years, the need for heritage preservation and reconstruction has become evi-
dent as many mature buildings face the risk of deterioration, damage or loss due to fac-
tors such as urban development, environmental weathering as well as outdated infras-
tructure. This urgency has created surges of significant interest to find sustainable meth-
ods of heritage preservation. The rise of emerging digital technologies has introduced
a multitude of innovative methods for storing, analysing, and showcasing building data.
Technologies such as 3D LiDAR scanning, and Building Information Modelling enable de-
tailed documentation and virtual exploration of heritage sites, while digital databases and
archives facilitate the easy access and use of historical records. This project will attempt
to address a new method of heritage preservation by using Gaussian Splatting in con-
junction with segmentation methods to create a visually accurate model while also incor-
porating semantic labels.

Files

License info not available