Automatic point cloud building envelope segmentation (auto-CuBES) using machine learning

Inventors

Maldonado, Bryan P. • HAYES, NOLAN W. • Hun, Diana

Assignees

UT Battelle LLC

Abstract

Modern retrofit construction practices use 3D point cloud data of the building envelope to obtain the as-built dimensions. However, manual segmentation by a trained professional is required to identify and measure window openings, door openings, and other architectural features, making the use of 3D point clouds labor-intensive. Automatic point Cloud Building Envelope Segmentation (Auto-CuBES) algorithms can significantly reduce the time spent during point cloud segmentation. The Auto-CuBES algorithm inputs a 3D point cloud generated by commonly available surveying equipment and outputs a dimensioned wire-frame model of the building envelope. By leveraging unsupervised machine learning methods in the Auto-CuBES methods facades, windows, and doors can be identified while keeping the number of calibration parameters low. Additionally, some embodiments of Auto-CuBES can generate a heat map of each facade indicating nonplanar characteristics that are valuable for optimization of connections used in overclad envelope retrofits.

Core Innovation

The invention produces a wireframe model of a building envelope from a point cloud representing the building envelope. The point cloud is generated by scanning the building envelope with a LiDAR system, and includes coordinates, light intensity, and pixel color. The data processing apparatus extracts a subset of points representing each individual facade and automatically segments each facade into facade sections for building envelope wireframe model generation.

For each facade, the system partitions facade sections and divides each facade section into sets of features representing interior, wall, and exterior features, with the features associated with certain points of the facade section. Outliers are removed from each facade section, and a building envelope wire frame model is produced from the segmented facade sections having outliers removed. The wireframe model accurately represents dimensions of the building envelope and is presented through a user interface.

In the method, each individual facade is partitioned into facade sections where each facade section includes at least one opening, and wall points and opening points are extracted from the facade sections. The method determines an off-of-planeness coefficient for connection design and removes outliers from wall points and opening points based on user-defined calibration parameters provided via a user interface. A wireframe model of the building envelope is then generated based on a combination of the wall points, the opening points, and the off-of-planeness coefficient.

Claims Coverage

The document provides two independent claims, a system claim and a method claim. Across these independent claims, the core inventive features are the LiDAR point cloud to building-envelope wireframe-model workflow, facade extraction and facade-section partitioning, point-set feature partitioning into wall/opening and interior/wall/exterior, removal of outliers using user-provided calibration parameters, and wireframe model generation with dimensions and an off-of-planeness coefficient for connection design.

Across the independent claims, the inventive coverage centers on automatically extracting facade-related point subsets from a LiDAR building-envelope point cloud, partitioning facades into facade sections, forming wall/opening and interior/wall/exterior point sets, removing outliers using user-defined calibration parameters, and generating a building-envelope wireframe model that accurately represents dimensions and incorporates an off-of-planeness coefficient for connection design.

Stated Advantages

Documented Applications