Automated 3D Building Envelope Recognition from Point Clouds for Energy Analysis

Existing buildings now represents the greatest opportunity to improve building energy efficiency. Building energy performance (BEP) analysis is becoming increasingly important because decision makers can have a better visualization of their building's performance and quickly make the solution for improving building energy efficiency and reducing environmental impacts. Nowadays, it is a common practice to obtain point clouds of existing buildings through using 3D laser scanning technology for as-is building modeling. However, current methods need manual process to convert the point clouds into a building information model which is very time consuming and labor intensive. Therefore, the objective of this paper is to introduce a methodology that automatically recognizes the major components of the building envelope as individual objects from point clouds for energy simulation. In this paper, methodology of recognizing different building envelope components is introduced. Additionally, a field test on a residential house was made to validate the proposed algorithm, and in the test results it shows the building envelope components can be recognize as individual objects. The simplistic geometric outlines of all the components are recognized and ready for energy analysis.