Related Experiment Video
Updated: May 5, 2026
14:08
Automated Midline Shift and Intracranial Pressure Estimation based on Brain CT Images
Published on: April 13, 2013
42.8K
Research on the Improved ICP Algorithm for LiDAR Point Cloud Registration
Honglei Yuan1, Guangyun Li2, Li Wang1
1Institute of Geospatial Information, Information Engineering University, Zhengzhou 450001, China.
Sensors (Basel, Switzerland)
|August 14, 2025
Summary
This study reveals that scanning incidence angle significantly impacts LiDAR point cloud quality. An improved weighted Iterative Closest Point (ICP) algorithm, accounting for this, enhances registration accuracy by up to 30%.
Area of Science:
- Geomatics Engineering
- Computer Vision
- Metrology
Background:
- Point cloud registration algorithms have mature theoretical frameworks but often neglect scanning quality's impact.
- LiDAR point cloud accuracy and density are highly variable, influenced by laser scanners and environmental factors.
- Scanning incidence angle is a critical error source, especially in short-range applications.
Purpose of the Study:
- To systematically investigate the relationship between scanning incidence angles and point cloud quality.
- To develop an improved point cloud registration algorithm that accounts for incidence-angle-dependent errors.
Main Methods:
- Proposed an incident-angle-dependent weighting function for point cloud observations.
- Developed an improved weighted Iterative Closest Point (ICP) registration algorithm incorporating the weighting function.
Main Results:
- The proposed weighted ICP algorithm demonstrated approximately 30% higher registration accuracy than traditional ICP.
- Achieved a 10% improvement in registration accuracy compared to Faro SCENE's proprietary solution.
- Quantified the significant impact of scanning incidence angle on point cloud registration outcomes.
Conclusions:
- Accounting for scanning incidence angle is crucial for accurate point cloud registration.
- The developed weighted ICP algorithm offers a significant improvement for engineering and industrial measurement applications.
- This research highlights the need to consider sensor-specific error sources in point cloud processing.
