基于虚拟力场的海洋涡旋空间显著形状提取算法及应用

Automatic detection of significant spatial patterns of ocean eddies based on virtual force field and its application

基于海洋涡旋表面形态特征和“虚拟力场的形状提取方法”，提出了一种海洋遥感图像涡旋空间显著形状提取自动检测方法。首先采用常规检测算子对海洋遥感影像进行涡旋边缘的提取，并对其进行矢量化，继之在涡旋边缘进行参数化的基础上分析计算作用于边缘曲线之间的吸引力和融化力，最后采用一定的优化策略经过逐次迭代自动提取海洋涡旋的显著形状，并以多景不同来源的遥感影像为例，进行了各种海洋涡旋空间形状定量提取的实例研究，验证了该方法对于形态非常复杂的涡旋识别与提取的有效性。

An approach to automatic detection of significant spatial patterns of ocean eddies was proposed on the basis of the study of surface characteristics of ocean eddies and a shape extracting method based on the virtual force field. Firstly, eddy edges were detected from original satellite images using traditional detect algorithm. And then after vectorization of eddy edges, attraction force and fusion force between edge curves were analyzed and calculated. After that, significant spatial patterns of ocean eddies were detected automatically using iterative repetition followed by optimized rule. Finally, spatial pattern auto-detection of different types of ocean eddies was extracted using satellite images. This study verified that it is an effective way to identify and detect ocean eddies with complex form.