To overcome the shortcomings of the traditional manual detection of underwater targets in side-scan sonar (SSS) images, a real-time automatic target recognition (ATR) method is proposed in this paper. This method consists of image preprocessing, sampling, ATR by integration of the transformer module and YOLOv5s (that is, TR–YOLOv5s), and target localization. By considering the target-sparse and feature-barren characteristics of SSS images, a novel TR–YOLOv5s network and a down-sampling principle are put forward, and the attention mechanism is introduced in the method to meet the requirements of accuracy and efficiency for underwater target recognition. Experiments verified the proposed method achieved 85.6% mean average precision (mAP) and 87.8% macro-F₂ score, and brought 12.5% and 10.6% gains compared with the YOLOv5s network trained from scratch, and had the real-time recognition speed of about 0.068 s per image.

The limitations of underwater acoustic communications mean that the side-scan sonar data of an autonomous underwater vehicle (AUV) cannot be transmitted back and processed in real time, which means that targets cannot be detected in real time. To address the problem, this paper proposes an autonomous underwater vehicle-based side-scan sonar real-time detection method for underwater targets. First, the paper describes the system and operation of real-time underwater-target detection by the side-scan sonar mounted on the autonomous underwater vehicle. Next, it proposes a real-time processing method for side-scan sonar data, method for constructing a deep-learning-based underwater-target detection model, and real-time method for underwater-target detection based on navigation strip images, which, together, solve the three key technical problems of real-time data processing, deep-learning-based detection model construction, and real-time target detection based on the autonomous underwater vehicle. Finally, through sea-based experiments, the effectiveness of the proposed methods is evaluated, providing a new solution for the autonomous underwater vehicle-based side-scan sonar real-time detection of underwater targets.

Side-scan sonar (SSS) is used for obtaining high-resolution seabed images, but with low position accuracy without using the Ultra Short Base Line (USBL) or Short Base Line (SBL). Multibeam echo sounder (MBES), which can simultaneously obtain high-accuracy seabed topography as well as seabed images with low resolution in deep water. Based on the complementarity of SSS and MBES data, this paper proposes a new method for acquiring high-resolution seabed topography and surface details that are difficult to obtain using MBES or SSS alone. Firstly, according to the common seabed features presented in both images, the Speeded-Up Robust Features (SURF) algorithm, with the constraint of image geographic coordinates, is adopted for initial image matching. Secondly, to further improve the matching performance, a template matching strategy using the dense local self-similarity (DLSS) descriptor is adopted according to the self-similarities within these two images. Next, the random sample consensus (RANSAC) algorithm is used for removing the mismatches and the SSS backscatter image geographic coordinates are rectified by the transformation model established based on the correct matched points. Finally, the superimposition of this rectified SSS backscatter image on MBES seabed topography is performed and the high-resolution and high-accuracy seabed topography and surface details can be obtained.

To meet the needs of coastline efficient extraction and dynamic monitoring, this paper proposes a new method for coastline extraction by combining the tidal level and the digital elevation model (DEM) of the coastal zone from tilt photography. Firstly, the DEM of coastal zone was obtained by using unmanned aerial vehicle (UAV) tilt photography; at the same time, the accuracy of aerial triangulation(AT) is improved referencing to the constraint of water boundary points, and then the mean high water spring tide was obtained by combining tidal harmonic analysis and Global Navigation Satellite System (GNSS) tidal level. Finally, the coastline and the dynamic water-surface line are extracted from the DEM of the coastal zone by tracking the contour lines with the elevation of the mean high water springs (MHWS) and the instantaneous sea-surface elevation, respectively. The experiments carried out in the coastal zones of Liaoning Province, China, proved the proposed method and achieved better than 0.2 m of horizontal position accuracy and 0.1 m of the vertical accuracy.

To reduce the size and cost of an integrated infrared (IR) and green airborne LiDAR bathymetry (ALB) system, and improve the accuracy of the green ALB system, this study proposes a method to accurately determine water surface and water bottom heights using a single green laser corrected by the near water surface penetration (NWSP) model. The factors that influence the NWSP of green laser are likewise analyzed. In addition, an NWSP modeling method is proposed to determine the relationship between NWSP and the suspended sediment concentration (SSC) of the surface layer, scanning angle of a laser beam and sensor height. The water surface and water bottom height models are deduced by considering NWSP and using only green laser based on the measurement principle of the IR laser and green laser, as well as employing the relationship between NWSP and the time delay of the surface return of the green laser. Lastly, these methods and models are applied to a practical ALB measurement. Standard deviations of 3.0, 5.3, and 1.3 cm are obtained by the NWSP, water-surface height, and water-bottom height models, respectively. Several beneficial conclusions and recommendations are drawn through the experiments and discussions.

To solve the shortage of training samples when using deep learning to detect shipwrecks, a comprehensive sample augmentation method is proposed. The method fully considers the imaging mechanism and environment of side-scan sonar (SSS), such as acoustic emission and reception, waterbody, target reflection, and seafloor background, and generates diverse and representative shipwreck samples from five aspects of target diversity, target texture, imaging resolution, equipment and environmental noises, and background through a series of novel sample augmentation methods. Under the condition of zero real SSS samples, a detection model of YOLOv5s was established with these amplified samples and achieved a mean average precision (MAP) better than 96% for real SSS data detection.