Common methods of ocean remote sensing and seafloor surveying are mainly carried out by airborne and spaceborne hyperspectral imagers. However, the water column hinders the propagation of sunlight to deeper areas, thus limiting the scope of observation. As an emerging technology, underwater hyperspectral imaging (UHI) is an extension of hyperspectral imaging technology in air conditions, and is undergoing rapid development for applications in shallow and deep-sea environments. It is a close-range, high-resolution approach for detecting and mapping the seafloor. In this paper, we focus on the concepts of UHI technology, covering imaging systems and the correction methods of eliminating the water column’s influence. The current applications of UHI, such as deep-sea mineral exploration, benthic habitat mapping, and underwater archaeology, are highlighted to show the potential of this technology. This review can provide an introduction and overview for those working in the field and offer a reference for those searching for literature on UHI technology.

Real-time safety assessment of the complex dynamic systems in nonstationary environments is of great significance for avoiding the potential hazards. In this case, the update procedure with high assessment accuracy and training speed is crucial and meaningful in the dynamic streaming setting. Generally, the performance of most online learning approaches will be negatively affected by limited annotated samples in such a setting. Moreover, the time cost of advanced conventional methods with retaining procedures is relatively high, constraining their practicality. In this article, a novel online active broad learning approach, termed OABL, is proposed. In detail, the effectiveness of the broad learning system in the framework of online active learning is first revealed and verified. A reasonable dynamic asymmetric query strategy is then designed with a limited annotation budget to actively annotate the relatively valuable samples, which is beneficial to mitigating the negative effects of class imbalance. In this context, the advantage of the human-in-the-loop characteristic is also effectively used to control the evolution direction of the learner during the incremental update, which makes it better able to adapt to complex and nonstationary environments. Several related experiments are conducted with the realistic data of JiaoLong deep-sea manned submersible. Results show the effectiveness and practicality of the proposal compared with the existing advanced approaches.

This article is concerned with the data-driven health state assessment task with unknown scenarios. Unknown scenarios are inaccessible in the stage of model training but they can appear unexpectedly during the running stage. A new problem called the within-class distribution mismatch is raised by assuming that unknown scenarios still belong to known classes. To tackle this challenging problem, a novel active incremental learning scheme with a classifier and an out-of-distribution (OOD) detector is proposed. A k -fast incremental support vector data description ( k -FISVDD) model is put forward as the OOD detector to recognize distribution mismatch samples online for label annotation. Specifically, it integrates the clustering algorithm to build local support vector sets, based on which an active query strategy is developed. An incremental learning mechanism is also designed to reduce the labeling cost. Then, the new labeled data can simultaneously refine the classifier and OOD detector. Two cases, including a bearing benchmark dataset and the operation data of a practical deep-sea manned submersible, are studied to demonstrate the effectiveness of the proposed scheme.

The processing of underwater images can vastly ease the difficulty of underwater robots’ tasks and promote ocean exploration development. This paper proposes a fast and efficient underwater image enhancement model based on conditional GAN with good generalization ability using aggregation strategies and concatenate operations to take full advantage of the limited hierarchical features. A sequential network can avoid frequently visiting additional nodes, which is beneficial for speeding up inference and reducing memory consumption. Through the structural re-parameterization approach, we design a dual residual block (DRB) and accordingly construct a hierarchical attention encoder (HAE), which can extract sufficient feature and texture information from different levels of an image, and with 11.52% promotion in GFLOPs. Extensive experiments were carried out on real and artificially synthesized benchmark underwater image datasets, and qualitative and quantitative comparisons with state-of-the-art methods were implemented. The results show that our model produces better images, and has good generalization ability and real-time performance, which is more conducive to the practical application of underwater robot tasks.

Considering the inherent unknown risks in the operation of complex dynamic system, it is very important to assess its safety accurately. Within the framework of human-in-the-loop, we propose a novel evidential group interaction-based (EGIB) safety assessment approach based on the evidence group interaction. With the analysis of safety assessment levels (SALs), the basic belief assignment (BBA) distribution of multiple safety indicators can be obtained. Fuzzy ranking technology is used to collect the individual’s cognition of these indicators. Using Shapley function in the framework of fuzzy measure and considering the cooperation effect, the peer importance is modeled. The collective group opinions of these indicators are obtained after the convergence process. The fusion procedure with weighted evidence combination is used to identify the safety state. To illustrate the effectiveness and practicability of the proposed method, we carry out simulation using real data obtained from electrical and observation communication system of deep-sea manned submersible, and comparsion is conducted between our approach and the baseline methods. Results show that the proposed approach can be applied to solve the safety assessment problems of complex dynamic systems in practice.

Zhang, Y.; Ding, Z.; Wang, Y.; Zhao, Q., and Qin, S., 0000. Determining the damage mechanisms for buoyancy materials of deep-sea manned submersibles. Deep-sea manned submersibles are important platforms for ocean exploration, and their structural reliability is critical to the personal safety of the crew. In particular, the buoyancy materials of deep-sea manned submersibles are vital parts because they provide the necessary buoyancy for submersible operations. This paper first describes the overall damage of the buoyancy materials used in the deep-sea manned submersible Jiaolong that accumulated over a large number of research expeditions during years of services. The causes of the damage are analyzed by computational fluid dynamics (CFD) methods. Because Jiaolong's buoyancy materials are secured to its structural frame by bolts, the external forces that act on the submersible are concentrated on these bolts. Those external forces include the wave forces that act on the submersible when it floats on the sea surface and the buoyancy material shrinkage that occurs because of changes in environmental pressure and temperature during submarine operations. Finally, several suggestions are provided for improving the design and operation of deep-sea submersibles.