Octree-based segmentation for terrestrial LiDAR point cloud data in industrial applications

Published Date
ISPRS Journal of Photogrammetry and Remote Sensing
March 2016, Vol.113:59–74, doi:10.1016/j.isprsjprs.2016.01.001

• Author 

• Yun-Ting Su ^a,,
• James Bethel ^a
• Shuowen Hu ^b

• aGeomatics Engineering, School of Civil Engineering, Purdue University, 550 Stadium Mall Drive, West Lafayette, IN 47907, United States
• bU.S. Army Research Laboratory, 2800 Powder Mill Road, Adelphi, MD 20783, United States

Received 18 April 2015. Revised 25 December 2015. Accepted 4 January 2016. Available online 19 January 2016. 

Abstract

Automated and efficient algorithms to perform segmentation of terrestrial LiDAR data is critical for exploitation of 3D point clouds, where the ultimate goal is CAD modeling of the segmented data. In this work, a novel segmentation technique is proposed, starting with octree decomposition to recursively divide the scene into octants or voxels, followed by a novel split and merge framework that uses graph theory and a series of connectivity analyses to intelligently merge components into larger connected components. The connectivity analysis, based on a combination of proximity, orientation, and curvature connectivity criteria, is designed for the segmentation of pipes, vessels, and walls from terrestrial LiDAR data of piping systems at industrial sites, such as oil refineries, chemical plants, and steel mills. The proposed segmentation method is exercised on two terrestrial LiDAR datasets of a steel mill and a chemical plant, demonstrating its ability to correctly reassemble and segregate features of interest.

Keywords

Terrestrial LiDAR

Point cloud segmentation

Split and merge

Piping systems

Cylinder detection

---

Fig. 1.

Fig. 2.

Fig. 3.

Fig. 4.

Fig. 5.

Fig. 6.

Fig. 7.

Fig. 8.

Fig. 9.

Fig. 10.

Fig. 11.

Fig. 12.

Fig. 13.

Fig. 14.

Fig. 15.

Fig. 16.

Fig. 17.

Fig. 18.

 Table 1

Table 1.

 Table 2

Table 2.

 Table 3

Table 3.

 Table 4

Table 4.

• ⁎ 
  Corresponding author.

Copyright © 2016 International Society for Photogrammetry and Remote Sensing, Inc. (ISPRS). Published by Elsevier B.V. All rights reserved.

For further details log on website :
http://www.sciencedirect.com/science/article/pii/S0924271616000022