Robust Underwater Laser Line Scanning System via Advanced Laser Line Extraction
Underwater laser line scanning plays a significant role in achieving precise 3-D reconstructions in challenging underwater environments. However, underwater refraction and scattering pose challenges in accurately extracting laser lines from the images. Additionally, the correction of intrinsic errors in the rotating scanning system is necessary. This article presents a novel underwater rotating laser line scanning system designed for high-precision 3-D reconstruction. Our system features an enhanced hand-eye calibration process, aimed at improving the accuracy of registration between the camera coordinate system and the rotation axis coordinate system by employing a multisampling approach. To address the complexities of underwater images, we propose a multiconstraint segmentation method for extracting laser lines, aimed at mitigating extraction errors arising from fluctuating light intensities and scattering in underwater environments. Additionally, our method integrates recent advancements in techniques for refraction correction, effectively mitigating the impact of underwater refraction and further enhancing the precision of our system. Experimental results demonstrate its robustness in extracting laser lines across diverse environments and delivering superior high-precision 3-D reconstructions.
Authors
- Hui Yu (ORCID: https://orcid.org/0000-0002-7655-9228)
- Junyu Dong (ORCID: https://orcid.org/0000-0001-7012-2087)
- Zhengwu Shi (ORCID: https://orcid.org/0000-0001-8808-0460)
- Hao Fan (ORCID: https://orcid.org/0000-0001-9133-8135)
- Zhihao Zhu (ORCID: https://orcid.org/0009-0001-4165-8250)
- Jiawei Wang (ORCID: https://orcid.org/0009-0008-8548-0061)
- Haitao Ren (ORCID: https://orcid.org/0009-0008-8111-0768)
Institutions
- Ocean University of China (CN)
- University of Portsmouth (GB)
- University of Glasgow (GB)
Publication Details
- Journal
- IEEE Transactions on Instrumentation and Measurement
- Published
- 2025-01-01
- DOI
- https://doi.org/10.1109/tim.2025.3551829
- Citations
- 1
- Primary Topic
- Advanced Optical Sensing Technologies
- Type
- article
- Field-Weighted Citation Impact
- 1.19
Funders
- National Natural Science Foundation of China