为满足船舶在复杂海洋环境下导航需求，在船舶导航系统设计中应用视频处理技术。采集航行视频流，并进行预处理后，运用帧间差分法构建自适应背景模型，划分海上目标前景、背景；并采用主成分分析法从前、背景划分结果中实时检测海上运动目标，精准获取目标位置、角度及运动趋势等关键信息，采用人工势场算法依据目标定位结果构建吸引与排斥势场，通过合力计算动态生成导航路径，成功规避多种类型障碍物，完成船舶路径导航。实验结果显示，视频处理技术划分的前景、背景图像边界清晰，且前景目标轮廓完整，背景图像未出现明显静态偏移或动态模糊，不存在误判情况，可精准获取目标位置、角度及运动趋势信息，有效导航出规避暗礁以及移动船舶的平滑路径。
To meet the navigation needs of ships in complex marine environments, video processing technology is applied in the design of ship navigation systems. After collecting and preprocessing the navigation video stream, an adaptive background model is constructed using the inter-frame difference method to distinguish between the foreground and background of maritime targets. The principal component analysis method is employed to detect moving targets at sea in real-time from the foreground and background segmentation results, accurately obtaining key information such as target position, angle, and motion trend. The artificial potential field algorithm is used to construct attractive and repulsive potential fields based on the target positioning results. The navigation path is dynamically generated through the calculation of resultant forces, successfully avoiding various types of obstacles and completing the ship's path navigation. Experimental results show that the foreground and background images segmented by video processing technology have clear boundaries, and the foreground target contours are complete. The background images do not exhibit significant static shifts or dynamic blurring, eliminating misjudgments. The system can accurately obtain information about target position, angle, and motion trend, effectively navigating a smooth path that avoids reefs and moving ships.
2025,47(20): 195-199 收稿日期：2025-4-30
DOI：10.3404/j.issn.1672-7649.2025.20.031
分类号：U665.13
作者简介：李播阳(1984-),女,硕士,讲师,研究方向为视频剪辑及信息可视化
参考文献：
[1] 何世全, 赵东, 王蕊, 等. 基于多传感器融合的无人艇自主导航系统的设计与实现[J]. 船舶工程, 2025, 47(5): 87-93.
HE S Q, ZHAO D, WANG R, et al. Design and implementation of autonomous navigation system for unmanned boat based on multi-sensor fusion[J]. Ship Engineering, 2025, 47(5): 87-93.
[2] 李元奎, 索基源, 于东冶, 等. A*与NSGA-II融合的船舶气象航线多目标规划方法[J]. 中国舰船研究, 2025, 20(3): 288-295.
LI Y K, SUO J Y, YU D Y, et al. Multi-objective programming method for ship weather routing based on fusion of A* and NSGA-Ⅱ[J]. Chinese Journal of Ship Research, 2025, 20(3): 288-295.
[3] 叶嘉宁, 谢博祎, 孙俊锋, 等. 面向导航服务的水网地区船舶航线规划[J]. 中国航海, 2024, 47(4): 60-65+72.
[4] 黄国良, 周毅, 郑坤, 等. 基于改进蚁群算法的全局船舶路径规划方法[J]. 船海工程, 2023, 52(2): 97-101,136.
HUANG G L, ZHOU Y, ZHENG K, et al. Ship path planning and collision avoidance based on improved ant colony algorithm[J]. Ship & Ocean Engineering, 2023, 52(2): 97-101,136.
[5] 吴伟, 王宁, 王元元, 等. 多特征融合的无人船视觉目标跟踪[J]. 大连海事大学学报, 2023, 49(4): 37-45+56.
[6] 潘云伟, 李敏, 曾祥光, 等. 基于人工势场和改进强化学习的自主式水下潜航器避障和航迹规划[J]. 兵工学报, 2025, 46(4): 72-83.
[7] 王群朋, 李龙浩, 关宏旭, 等. 基于三维势场模型的船舶局部路径规划方法[J]. 中国舰船研究, 2025, 20(1): 135-146.
WANG Q P, LI L H, GUAN H X, et al. Ship local path planning method based on three-dimensional potential field model[J]. Chinese Journal of Ship Research, 2025, 20(1): 135-146.
