Abstract
Indoor environments are challenging for multisensor calibrations. Traditional calibration methods use the target structure for camera and LiDAR calibration. Those approaches not only require pre-processed data and offline calculations, but also face challenges in low-light and object-occluded indoor environments. We proposed an automatic calibration method using trajectory constraints on the LiDAR-Camera. The proposed method first obtains the accurate LiDAR trajectory by the LiDAR-SLAM (LIO-SAM) algorithm. At the same time, the problem of visual SLAM trajectory drift in the indoor environment is improved by graphical optimization using the rigid relative position invariance between sensors during displacement. Thus, extrinsic calibration is achieved by using the relative relationship between sensor trajectories. This method has higher robustness than the target-based calibration methods. The experimental results show that our algorithm has higher accuracy than the target-based calibration in the underground environment. The rotation root-mean-square error (RMSE) improves from 6.637° to 0.564°, and the translation RMSE improves from 0.197 to 0.082 m.
| Original language | English |
|---|---|
| Article number | 8501707 |
| Journal | IEEE Transactions on Instrumentation and Measurement |
| Volume | 72 |
| DOIs | |
| State | Published - 2023 |
| Externally published | Yes |
Keywords
- Low-cost platform
- mobile laser scanning
- multisensor calibration
- point cloud
- simultaneous localization and mapping (SLAM)
Fingerprint
Dive into the research topics of 'An Overlap-Free Calibration Method for LiDAR-Camera Platforms Based on Environmental Perception'. Together they form a unique fingerprint.Cite this
- APA
- Author
- BIBTEX
- Harvard
- Standard
- RIS
- Vancouver