Determining the appropriate number and position of waypoints on a great circle route (GCR) helps to shorten the sailing distance, reduce the number of course changes, and well-approximate the GCR through a small number of rhumb line (RL) legs. In this study, a genetic algorithm-based method (i.e., the GA method) is proposed to optimize the positions of waypoints on the GCR when the number of waypoints is given. Furthermore, a fuzzy logic-based evaluation method for the number of waypoints (i.e., the FL method) is proposed to judge whether to add a new waypoint or stop the process by using the non-fixed values while considering both the number of waypoints and the remaining benefit of the GCR. According to the example demonstration results, the two methods proposed in this study can well-determine the number and position of waypoints and provide effective support for ocean route planning.

3D object perception in maritime is a promising means to enhance the ability of the autonomous navigation. However, 3D automatic reconstruction of dynamic targets remains a challenging task in the 3D object perception systems. In this paper, we present an innovative GAN-based deep learning neural network for 3D ship reconstruction. The method is able to reconstruct 3D ship models from single view images or videos using shape prior knowledge and a vessel generative adversarial network (Vessel-GAN), which can effectively extract the ship shape features from the single-view images or videos. Furthermore, a 3D feature network optimization model is proposed for loop optimization of 3D ship shape perception. In order to validate the effectiveness of the proposed method, a high resolution single ship dataset (HSSD) was created, and the performance of the proposed method was evaluated against the existing methods, The experimental results show that the Vessel Segment Network (VSNet) can improve the accuracy of the HSSD dataset by 2.5% in the single ship segmentation tasks. In addition, Vessel-GAN was demonstrated to be able to generate multi-view ship images with an impressive FID value of 5.82. Moreover, our model can improve the accuracy of 3D ship modeling by 8.9% on the ShapeNet dataset. This innovative approach promises to enable 3D perception of the marine environment for autonomous ship.

Mastering the real-time dynamics of near-field sea ice is the primary condition to guaranteeing the navigation safety of autonomous ships in the polar region. In this study, a radar image fusion process combining marine radar and ice radar is proposed, which can effectively solve the problems of redundant information and spatial registration during image fusion. Then, using the fused radar images, this study proposes a set of near-field sea ice risk assessment and warning processes applicable to both low- and high-sea-ice-concentration situations. The sea ice risk indexes in these two situations are constructed by using four variables: sea ice area, sea ice grayscale, distance between sea ice and the own-ship, and relative bearing of sea ice and the own-ship. Finally, visualization processing is carried out according to the size of the risk index values of each piece of sea ice to achieve a better near-field sea ice risk assessment and warning effect. According to the example demonstration results, through the radar image fusion process and the set of near-field sea ice risk assessment and warning processes proposed in this study, the sea ice risk distribution in the near-field area of the ship can be well obtained, which provides effective support for the assisted decision-making of autonomous navigation in the polar region.