Oil spill detection method for marine radar based on multi-dimensional features and improved MShOA
Abstract
溢油严重威胁海洋环境与生态系统健康，高效的监测技术是减损溢油事故、制定应急策略的关键。针对传统航海雷达监测方法易受噪声干扰的问题，本研究提出融合多维度特征聚类与改进螳螂虾优化算法（mantis shrimp optimization algorithm，MShOA）混合策略的监测方法，通过生成多维特征、K-Means聚类及改进的MShOA智能优化搜索实现油膜高效分割。改进的MShOA引入动态惯性权重与高斯扰动机制，提升优化搜索能力。实验表明，该方法在航海雷达数据溢油检测方面具有较强的鲁棒性，可以为海上溢油监测提供可靠方案。
Fields of Study & Subjects
Oil Spill Detection and Mitigation
Maritime Navigation and Safety
Fire Detection and Safety Systems
Publication details
- Journal / source
- Marine Environmental Science
- Publisher
- China Science Publishing & Media Ltd.
- Volume
- 45
- Issue
- 3
- Pages
- 417–423
- Publication date
- 30 January 2026
- Language
- EN