Chinese

Hybrid-VINS: Underwater Tightly-Coupled Hybrid Visual Inertial Dense SLAM for AUV

Published in IEEE Transactions on Industial Electronics (major revision), 2024

1) An underwater tightly-coupled hybrid visual inertial dense SLAM framework, named Hybrid-VINS, is proposed, which is more suitable for underwater scenarios. To the best of our knowledge, this is the first underwater SLAM system to utilize active vision information to assist passive vision. 2) The self-designed structured light system is used to correct the depth measurement of some features during passive vision initialization and tracking, which improves the localization accuracy. In addition, the introduction of the structured light system information realizes the VINS dense mapping, which is very rare underwater. 3) A more robust hybrid vision-aided loop closure detection algorithm is proposed to overcome the inaccuracy of purely passive vision loop factor. 4) The underwater autonomous hybrid vision system is developed in an underwater simulation environment and the real underwater world, respectively, to collect multiple datasets to validate the performance of Hybrid-VINS.

Structured Light-Based Underwater Collision-Free Navigation and Dense Mapping System for Refined Exploration in Unknown Dark Environments

Published in IEEE Transactions on Systems, Man, and Cybernetics: Systems, 2024

1) A more adaptable 3-D dense mapping robotic system based on self-designed scanning BSL, named ROVScanner, is developed for refined exploration, where the on-board design allows for autonomous mobility and operational capabilities. A more efficient underwater 3-D dense mapping algorithm fusing DVL, inertial measurement unit (IMU) and pressure sensor multifrequency information is proposed to realize dense mapping during robot motion. 2) An air–water two-stage underwater multisensor calibration method is presented. In particular, the extrinsic parameters between DVL and camera are innovatively calibrated using BSL based on graph optimization, enhancing robustness. 3) A framework of BSL-based collision-free navigation is presented to guarantee the safe movement of the system in unknown dark environments. To the best of our knowledge, this is the first work that can simultaneously realize autonomous collision-free navigation and dense mapping in dark underwater environments by utilizing active structured light vision.

Recommended citation: Y. Ou et al., "Structured Light-Based Underwater Collision-Free Navigation and Dense Mapping System for Refined Exploration in Unknown Dark Environments," in IEEE Transactions on Systems, Man, and Cybernetics: Systems, doi: 10.1109/TSMC.2024.3370917. http://ouyaming.github.io/files/2024-03-18-TSMC.pdf

Structured light vision based pipeline tracking and 3D reconstruction method for underwater vehicle

Published in IEEE Transactions on Intelligent Vehicles, 2024

1) A novel underwater pipeline positioning method based on dual-line laser SLV is proposed, which can simultaneously obtain the lateral deviation, height deviation and heading deviation of underwater vehicle and underwater pipeline under weak light water environment, providing the basis for underwater pipeline tracking. 2) By combining laser stripe image feature points, refracted underwater SLV model and Doppler Velocity Log (DVL) information, the tracking and dense 3D reconstruction of underwater pipeline are realized, which is difficult for existing underwater inspection methods. 3) By integrating the self-designed underwater SLV sensor with the underwater vehicle BlueROV, an underwater pipeline tracking and 3D reconstruction system is developed, and a series of planar and spatial pipeline experiments are carried out to verify its effectiveness.

Recommended citation: J. Fan, Y. Ou, X. Li, C. Zhou and Z. Hou, "Structured Light Vision Based Pipeline Tracking and 3D Reconstruction Method for Underwater Vehicle," in IEEE Transactions on Intelligent Vehicles, vol. 9, no. 2, pp. 3372-3383, Feb. 2024, doi: 10.1109/TIV.2023.3340737. http://ouyaming.github.io/files/2023-12-08-TIV.pdf

Water-MBSL: Underwater Movable Binocular Structured Light-Based High-Precision Dense Reconstruction Framework

Published in IEEE Transactions on Industrial Informatics, 2023

1) A more adaptable 3-D dense mapping robotic system based on self-designed scanning BSL, named ROVScanner, is developed for refined exploration, where the on-board design allows for autonomous mobility and operational capabilities. A more efficient underwater 3-D dense mapping algorithm fusing DVL, inertial measurement unit (IMU) and pressure sensor multifrequency information is proposed to realize dense mapping during robot motion. 2) An air–water two-stage underwater multisensor calibration method is presented. In particular, the extrinsic parameters between DVL and camera are innovatively calibrated using BSL based on graph optimization, enhancing robustness. 3) A framework of BSL-based collision-free navigation is presented to guarantee the safe movement of the system in unknown dark environments. To the best of our knowledge, this is the first work that can simultaneously realize autonomous collision-free navigation and dense mapping in dark underwater environments by utilizing active structured light vision.

Recommended citation: Y. Ou, J. Fan, C. Zhou, L. Cheng and M. Tan, "Water-MBSL: Underwater Movable Binocular Structured Light-Based High-Precision Dense Reconstruction Framework," in IEEE Transactions on Industrial Informatics, vol. 20, no. 4, pp. 6142-6154, April 2024, doi: 10.1109/TII.2023.3342899. http://ouyaming.github.io/files/2023-12-29-TII.pdf

Binocular structured light 3-D reconstruction system for low-light underwater environments: Design, modeling, and laser-based calibration

Published in IEEE Transactions on Instrumentation and Measurement, 2023

1) An underwater binocular structured light 3-D reconstruction system with the scanning laser is designed to realize the static high-precision scanning reconstruction of the low-light scene, which is suitable for underwater robot application, including grasping, surveying, and mapping. The obtained high-precision 3-D point clouds prove the effectiveness of our system. 2) Three models based on underwater refraction effects are systematically proposed, among which the multimedia binocular polar curve constraint model ensures accurate laser line matching, which is a relatively cutting-edge work. 3) A simple special calibration block is designed, and a new multiobjective laser-based calibration algorithm based on laser geometric constraints is proposed. The proposed method only needs one scanning data, which greatly simplifies the calibration process. More importantly, this method could achieve accurate calibration in the low-light underwater environment, which is hard for the checkerboard-based calibration method.

Recommended citation: Y. Ou, J. Fan, C. Zhou, S. Tian, L. Cheng and M. Tan, "Binocular Structured Light 3-D Reconstruction System for Low-Light Underwater Environments: Design, Modeling, and Laser-Based Calibration," in IEEE Transactions on Instrumentation and Measurement, vol. 72, pp. 1-14, 2023, Art no. 5010314, doi: 10.1109/TIM.2023.3261941. http://ouyaming.github.io/files/2023-03-27-TIM.pdf

Data Calibration Algorithm for Artificial Lateral Line Sensor of Robotic Fish on Improved LSTM

Published in 2021 40th Chinese Control Conference (CCC), 2021

1) A cantilever-type bionic lateral line sensor unit based on the piezoresistive effect is designed. This unit consists of three parts: a flow-facing plane, a horizontal cantilever, and a strain gauge. It can effectively simulate the lateral line system of real fish and achieve the purpose of speed measurement for robots. 2) On this basis, a precise hydrodynamic model of the unit is established to provide a theoretical basis for information conversion. Then, an improved LSTM-based data calibration algorithm is proposed, achieving accurate calibration of the sensor.

Recommended citation: Y. Ou, Z. Zhang, C. Zhou and B. Zhou, "Data Calibration Algorithm for Artificial Lateral Line Sensor of Robotic Fish on Improved LSTM," 2021 40th Chinese Control Conference (CCC), Shanghai, China, 2021, pp. 4308-4314, doi: 10.23919/CCC52363.2021.9549820. http://ouyaming.github.io/files/2021-07-26-CCC.pdf