Related Experiment Video
Updated: Dec 20, 2025
08:16
Collecting and Processing Drone-based Remotely Sensed Data for Use in Forest Recovery Monitoring
Published on: October 24, 2025
335
Method for extraction of airborne LiDAR point cloud buildings based on segmentation.
Maohua Liu1,2, Yue Shao2, Ruren Li2
1College of Land and Environment, Shenyang Agricultural University, Shenyang, China.
Plos One
|May 30, 2020
Summary
This study introduces an efficient airborne LiDAR building point cloud extraction method using Point Cloud Library (PCL) region growing and histograms. The novel approach reduces errors and improves building extraction accuracy compared to existing methods.
Area of Science:
- Geospatial technology
- Remote sensing
- Computer vision
Background:
- Airborne LiDAR is crucial for urban 3D modeling.
- Extracting building point clouds is a key but complex step.
- Existing methods often require multiple feature parameters.
Purpose of the Study:
- To propose a novel, efficient building point cloud extraction method.
- To reduce complexity and improve accuracy in LiDAR data processing.
- To effectively separate building point clouds from non-building data.
Main Methods:
- Utilizing Point Cloud Library (PCL) region growing for initial segmentation.
- Calculating local normal vectors and direction cosines for segmented clusters.
- Employing histogram analysis for final building point cloud separation.
Main Results:
- The proposed method was tested on airborne LiDAR data from Tokushima, Japan.
- Comparison with TerraSolid and K-means algorithms showed superior performance.
- Lower type I and II errors were achieved, indicating better extraction accuracy.
Conclusions:
- The combined PCL region growing and histogram method is effective for airborne LiDAR building extraction.
- This approach offers improved accuracy and efficiency over traditional methods.
- The technique contributes to more precise urban 3D modeling.
