Laser & Optoelectronics Progress, Volume. 57, Issue 20, 202802(2020)
Multi-Resolution 3D Reconstruction of Karst Caves Based on the Feature Line Extraction of 3D Laser Point Cloud
The surfaces of caves are complex and irregular. Many existing modeling methods are based on an overall-resolution 3D reconstruction. Although the overall model resolution is guaranteed, the efficiency of 3D reconstruction is substantially low, and the model file is too large, which makes it considerably difficult for the follow-up professional application. Therefore, a multi-resolution 3D reconstruction technology based on 3D laser point cloud feature line extraction was proposed for the karst cave. First, the enhanced geometric features of neighboring points were used to extract the eigenvalues and increased the normal vector angle as a basis for detecting the feature points. Second, the standard particle swarm optimization(SPSO) and fuzzy C-means clustering algorithms were used to realize the point cloud classification. Third, the broken line growth method was used to connect the feature points into the feature lines and project them onto 3D point cloud; finally, the classified point cloud was used for modeling as per different resolutions. The 3D reconstruction of the karst cave was realized with high precision, high quality, and high efficiency. The experimental results show that this method can improve the efficiency of the 3D reconstruction of karst caves, realize modeling according to different resolutions, reduce the amount of data after the 3D reconstruction, improve the efficiency of 3D reconstruction, and has high practical value.
Get Citation
Copy Citation Text
Hongqiang Bai, Yonghua Xia, Minglong Yang, Zhaoyong Li, De Huang. Multi-Resolution 3D Reconstruction of Karst Caves Based on the Feature Line Extraction of 3D Laser Point Cloud[J]. Laser & Optoelectronics Progress, 2020, 57(20): 202802
Category: Remote Sensing and Sensors
Received: Jan. 3, 2020
Accepted: Feb. 10, 2020
Published Online: Oct. 14, 2020
The Author Email: Xia Yonghua (617073761@qq.com), Yang Minglong (617073761@qq.com)