基于SAR图像的海洋中尺度涡弱纹理特征检测方法研究

Research on Weak Texture Feature Detection Method for Mesoscale Eddies in the Ocean Based on SAR Images

摘要: 海洋中尺度涡是一种典型的海洋中尺度现象之一，对水下声纳探测、全球气候系统、海洋生态环境以及渔业等具有严重影响，一直是海洋界研究的热点。全天时、全天候、高分辨率、宽幅SAR在海洋中尺度涡动态监测和特征参数反演精度方面具有明显优势，但是面临复杂海洋背景下中尺度涡反差信号弱、边缘特征检测难的问题。本文提出了多参数阈值法海洋中尺度涡边缘检测方法，实验结果表明该方法能够有效抑制背景噪声，显著提升SAR图像信杂比，实现海洋中尺度涡弱纹理边缘特征检测，为海洋中尺度涡精细化观测提供了新思路。

Abstract: Oceanic mesoscale eddies, as one of the typical oceanic mesoscale phenomena, have a significant impact on underwater sonar detection, the global climate system, marine ecological environment, and fisheries, and have always been a hot topic in oceanography research. All-day, all-weather, high-resolution, wide-swath Synthetic Aperture Radar (SAR) possesses obvious advantages in the dynamic monitoring of oceanic mesoscale eddies and the inversion accuracy of characteristic parameters. However, it faces the challenges of weak contrast signals and difficult edge feature detection of mesoscale eddies in complex marine backgrounds. This paper proposes a multi-parameter threshold method for edge detection of oceanic mesoscale eddies. Experimental results show that this method can effectively suppress background noise, significantly improve the signal-to-noise ratio of SAR images, and achieve weak texture edge feature detection of oceanic mesoscale eddies, providing a new approach for refined observation of oceanic mesoscale eddies.