Research and Application of Panoramic Visual Perception-Assisted Navigation Technology for Ships

Chiming Wang,Xiaocong Cai,Yanan Li,Runxuan Zhai,Rongjiong Wu,Shunzhi Zhu,Liangqing Guan,Zhiqiang Luo,Shengchao Zhang,Jianfeng Zhang

DOI: https://doi.org/10.3390/jmse12071042

IF: 2.744

2024-06-22

Journal of Marine Science and Engineering

Abstract:In response to challenges such as narrow visibility for ship navigators, limited field of view from a single camera, and complex maritime environments, this study proposes panoramic visual perception-assisted navigation technology. The approach includes introducing a region-of-interest search method based on SSIM and an elliptical weighted fusion method, culminating in the development of the ship panoramic visual stitching algorithm SSIM-EW. Additionally, the YOLOv8s model is improved by increasing the size of the detection head, introducing GhostNet, and replacing the regression loss function with the WIoU loss function, and a perception model yolov8-SGW for sea target detection is proposed. The experimental results demonstrate that the SSIM-EW algorithm achieves the highest PSNR indicator of 25.736, which can effectively reduce the stitching traces and significantly improve the stitching quality of panoramic images. Compared to the baseline model, the YOLOv8-SGW model shows improvements in the P, R, and mAP50 of 1.5%, 4.3%, and 2.3%, respectively, its mAP50 is significantly higher than that of other target detection models, and the detection ability of small targets at sea has been significantly improved. Implementing these algorithms in tugboat operations at ports enhances the fields of view of navigators, allowing for the identification of targets missed by AISs and radar systems, thus ensuring operational safety and advancing the level of vessel intelligence.

oceanography,engineering, marine, ocean

What problem does this paper attempt to address?

This paper mainly discusses the technology of ship panoramic visual perception assisted navigation to solve the navigation challenges faced by narrow vision of crew members, limited field of view of single camera, and complex marine environment. The study proposes a panoramic image stitching algorithm, SSIM-EW, which combines the SSIM (structural similarity index) region search method and the elliptical weighted fusion method, aiming to reduce stitching artifacts and improve the quality of panoramic images. In addition, the paper improves the YOLOv8s model by increasing the size of detection heads, introducing GhostNet, and replacing the regression loss function to enhance the performance of marine target detection, and proposes a perception model called yolov8-SGW. Experimental results show that the SSIM-EW algorithm performs best in terms of the PSNR index, while the yolov8-SGW model outperforms the baseline model in terms of P, R, and mAP50, especially significantly enhancing the detection capability of small targets.

These technologies applied to port tugboat operations can expand the crew's field of view, identify targets that may be missed by the AIS (Automatic Identification System) and radar systems, thereby improving navigation safety and promoting the level of ship intelligence. The main contribution of the paper is the development of a real-time environmental perception technology for building a ship visual-assisted perception system to support safer and smarter ship navigation.