Fiber optic sensors are expected to be an auxiliary measurement tool in the field of ocean observation due to their small size, easy networking, intrinsic immunity to electromagnetic interference, and many other advantages. In recent years, the research around fiber optic sensing technology in the field of physical ocean observation has received increasing attention from researchers. According to the different measurement objects and measurement principles, the focus is on the seawater temperature sensor based on fiber grating, seawater salinity sensor, seawater pressure sensor, seawater salinity sensor based on multi-core optical fiber, seawater salinity sensor based on Michelson and Mach-Zehnder interferometry, seawater pressure sensor based on Fabry-Perot cavity, fiber optic temperature salinity and depth sensors based on resonant and coupling technology, flow velocity sensor based on multi-core optical fiber and fiber optic turbulence sensors based on thermal dissipation and turbulent kinetic energy dissipation. The test methods and test results of various sensors are analyzed, and the advantages and disadvantages of the technologies are pointed out. Prospects for the application of fiber optic sensors in physical ocean observation are presented.

In a typical underwater acoustic target detection mission, we have to estimate the target number (N), perform source separation when $N>1$, and consequently predict the motion parameters such as fundamental frequency (F₀) from separated noises for each target. Although deep learning methods have been adopted in each task, their successes strongly depend on the feed-in features. In this paper, we evaluate several time-frequency features and propose a universal feature extraction strategy for object counting and F₀ estimation simultaneously, with a convolutional recurrent neural network (CRNN) as the backbone. On one hand, LOFAR and DEMON are feasible for low-speed and high-speed analysis, respectively, and are combined (LOFAR + DEMON) to cope with full-condition estimation. On the other hand, a comb filter (COMB) is designed and applied to the combined spectrum for harmonicity enhancement, which will be further streamed into the CRNN for prediction. Experiments show that (1) in the F₀ estimation task, feeding the filtered combined feature (LOFAR + DEMON + COMB) into the CRNN achieves an accuracy of 98% in the lake trial dataset, which is superior to LOFAR + COMB (83%) or DEMON + COMB (94%) alone, demonstrating that feature combination is plausible. (2) In a counting task, the prediction accuracy of the combined feature (LOFAR + DEMON, COMB included or excluded) is comparable to the state-of-the-art on simulation dataset and dominates the rest on the lake trial dataset, indicating that LOFAR + DEMON can be used as a common feature for both tasks. (3) The inclusion of COMB accelerates the convergence speed of the F₀ estimation task, however, it penalizes the counting task by a depression of 13% on average, partly due to the merging effects brought in by the broadband filtering of COMB.

Monitoring the field of operation of an underwater vehicle is crucial during missions near the sea floor. The forward-looking sonar is often the only available sensor for the observation of the ambient turbid water environment. Sonar image registration is not only a first step towards a panoramic mosaic but it also provides an initial motion parameter estimation for the vehicle self-localization and navigation. In this article, a peripheral mutual information (PMI) maximization method is proposed for the sonar image registration. Peripheral mutual information is inspired by regional mutual information (RMI) which makes use of the closed-form solution for the Shannon entropy by the assumption that the data vectors made of neighbouring pixels are normally distributed, an assumption that ignores correlations between the pixels in sonar images. To accommodate the fact that the neighbouring pixels show dependencies due to acoustic reverberation and dispersion, only the peripheral information in the neighbourhood of a pixel is used in peripheral mutual information for the calculation of the mutual information. Experiments show that the peripheral mutual information registration function is much smoother than that of regional mutual information. Further experiments on the two-dimensional forward-looking sonar image registration demonstrate the efficiency of peripheral mutual information.

Due to the unavailability of GPS signal, it is more urgent to develop the autonomous navigation capability for the underwater vehicles. In this paper, we summarize the development status of underwater SLAM (simultaneous localization and mapping) system. Different from the terrestrial or aerial SLAM that largely depends on the optical sensors, the underwater SLAM system mainly uses the acoustic sensors, i.e., sonars, to watch the environment. With respect to the general SLAM system, which is mainly composed of the front-end local data-association and the back-end global error adjustment, we briefly survey recent progress in sonar image registration and the loop closure detection. Furthermore, some heuristic problems are posed in the conclusion.

A local-autoencoding (LAE) method is proposed for the parameter estimation in a Hidden Potts-Markov random field model. Due to sampling cost, Markov chain Monte Carlo methods are rarely used in real-time applications. Like other heuristic methods, LAE is based on a conditional independence assumption. It adapts, however, the parameters in a block-by-block style with a simple Hebbian learning rule. Experiments with given label fields show that the LAE is able to converge in far less time than required for a scan. It is also possible to derive an estimate for LAE based on a Cramer-Rao bound that is similar to the classical maximum pseudolikelihood method. As a general algorithm, LAE can be used to estimate the parameters in anisotropic label fields. Furthermore, LAE is not limited to the classical Potts model and can be applied to other types of Potts models by simple label field transformations and straightforward learning rule extensions. Experimental results on image segmentations demonstrate the efficiency and generality of the LAE algorithm.

Robust and accurate estimation of position and attitude of a UUV (Unmanned Underwater Vehicle) from sonar scans is essential for simultaneous localization and mapping (SLAM). Both dead-reckoning based on the inertial navigation system and the motion parameter estimation based on the registration of the ultrasound scan sequence can contribute to the performance of the system. However, the rapidly-growing accumulated error tends to counteract the precise localization of the vehicle. In this paper, a method for loop-closure detection is proposed that adjusts the accumulated error for the underwater acoustic SLAM when the vehicle scans the underwater environment using an Mechanical Scanning Imaging Sonar (MSIS). Firstly, a similarity matrix for pairs of scans is constructed to represent the loop-closing tracks. In the registration step, two novel features, namely the intensity projection histograms and a polar gradient matrix, are extracted to calculate the translational and rotational parameters respectively. Secondly, the probability hypothesis density (PHD) filter is used to extract the possible loop-closure constraints from the similarity matrix, removing the random noise brought by accidental correlation and refining the concurrent loop-closing tracks resulted from long-range scanning. Lastly, the loop-closure constraints from the refined loop-closing tracks are fed into the GraphSLAM system to adjust the pose of each scan by constraint optimization. Experiments on the MSIS sonar images collected in structured and unstructured underwater environments validate the effectiveness of the proposed loop-closure detection method.