Seabed fishing depends on humans in common, for instance, the sea cucumber, sea urchin, and scallop fishing, which is always a very dangerous task. Considering the underwater complex environment conditions such as low temperature, dim vision, and high pressure, collecting the marine products using underwater robots is commonly regarded as a feasible solution. The key technique of the underwater robot development is to detect and locate the main target from underwater vision. This research is based on the deep convolutional neural network (CNN) to realize the target recognition from underwater vision. The RPN (Region Proposal Network) is used to optimize the feature extraction capability. Deep learning dataset is prepared using an underwater video obtained from a sea cucumber fishing ROV (Remote Operated Vehicle). The inspiration of the network structure and the improvements come from the Faster RCNN and Hypernet method, and for the underwater dataset, the method proposed in this paper shows a good performance of recall and object detection accuracy. The detection runs with a speed of 17 fps on a GPU, which is applicable to be used for real-time processing.

Underwater SLAM encounters challenges in complex environments due to suspended particles, underwater blur, and light and color attenuation. These factors make underwater features less distinct than those in surface images. Moreover, underwater visual features are often unstructured, together with visual failures and low visibility. To address these challenges, a multi-sensor system is a promising solution. In this paper, we introduce a multi-sensor fusion underwater SLAM method, integrating stereo vision, multi-beam imaging sonar, and IMU data. Our system comprises a visual-inertial subsystem and an acoustic-inertial subsystem. These subsystems collaborate when common features are detected. If one subsystem fails, the other can function independently. The visual-inertial subsystem uses depth information from imaging sonar to optimize error correction during feature tracking. Furthermore, we have optimized the initialization process by matching visual and sonar images and introduced a novel method for depth estimation from sonar images. This dual-sensor strategy improves the system’s robustness and adaptability to diverse challenging underwater conditions. Through experiments, we have demonstrated the excellent performance of our algorithm.

In the realm of ocean engineering and maintenance of subsea structures, accurate underwater distance quantification plays a crucial role. However, the precision of such measurements is often compromised in underwater environments due to backward scattering and feature degradation, adversely affecting the accuracy of visual techniques. Addressing this challenge, our study introduces a groundbreaking method for underwater object measurement, innovatively combining image sonar with stereo vision. This approach aims to supplement the gaps in underwater visual feature detection with sonar data while leveraging the distance information from sonar for enhanced visual matching. Our methodology seamlessly integrates sonar data into the Semi-Global Block Matching (SGBM) algorithm used in stereo vision. This integration involves introducing a novel sonar-based cost term and refining the cost aggregation process, thereby both elevating the precision in depth estimations and enriching the texture details within the depth maps. This represents a substantial enhancement over existing methodologies, particularly in the texture augmentation of depth maps tailored for subaquatic environments. Through extensive comparative analyses, our approach demonstrates a substantial reduction in measurement errors by 1.6%, showing significant promise in challenging underwater scenarios. The adaptability and accuracy of our algorithm in generating detailed depth maps make it particularly relevant for underwater infrastructure maintenance, exploration, and inspection.

The advancement of intelligent technology has propelled the development of smart unmanned vessels into a new phase. To address the urgent demands of current smart ship development, this paper develops a comprehensive ship digital twin system based on a virtual-real mapping algorithm, focusing on the fundamental elements of digital twin model construction. Using the smart unmanned experimental ship Dolphin 1 as a prototype, a digital twin virtual model is proposed. This system leverages real-time internal and external data from the entire vessel to track its navigational status, performance indicators, sailing trends, and surrounding flow field information, offering coordinated “human-machine” navigation assistance. Based on historical data collected from the vessel's long-term navigation, a real-time precise prediction of the vessel's navigational state and hydrodynamic performance is conducted using physics-informed neural network algorithm. This establishes a self-learning iterative virtual-physical mapping model that enables autonomous updates and evolution. As the real navigation data of the vessel continuously update, the virtual model can more accurately simulate the vessel's state in real time. The proposed digital twin model has been tested through sea trials under real sea conditions, demonstrating its high accuracy, robustness, and potential for enhancing navigational safety and efficiency. This system marks a significant step forward in the integration of digital twin technology with maritime navigation, providing a valuable tool for the future development of smart shipping.

In real-world underwater environment, exploration of seabed resources, underwater archaeology, and underwater fishing rely on a variety of sensors, vision sensor is the most important one due to its high information content, non-intrusive, and passive nature. However, wavelength-dependent light attenuation and back-scattering result in color distortion and haze effect, which degrade the visibility of images. To address this problem, firstly, we proposed an unsupervised generative adversarial network (GAN) for generating realistic underwater images (color distortion and haze effect) from in-air image and depth map pairs based on improved underwater imaging model. Secondly, U-Net, which is trained efficiently using synthetic underwater dataset, is adopted for color restoration and dehazing. Our model directly reconstructs underwater clear images using end-to-end autoencoder networks, while maintaining scene content structural similarity. The results obtained by our method were compared with existing methods qualitatively and quantitatively. Experimental results obtained by the proposed model demonstrate well performance on open real-world underwater datasets, and the processing speed can reach up to 125FPS running on one NVIDIA 1060 GPU. Source code, sample datasets are made publicly available at https://github.com/infrontofme/UWGAN_UIE.