论文部分内容阅读
在SAR强度影像中,包括海洋溢油在内的许多海洋现象呈现为暗斑。为从诸多暗斑中辨识海洋溢油,需要在SAR影像中提取暗斑的几何和统计分布特征,以此作为进一步分类(辨识)海洋溢油的依据,将基于几何划分技术的区域分割方法应用于SAR影像暗斑特征提取。首先建立高分辨率SAR影像暗斑或然率模型,然后利用最大化期望值和M-H算法实现其几何及统计分布特征参数提取。实验结果表明,该方法不仅可以精准提取暗斑的几何形状,同时还能有效估计其统计分布参数。
In SAR images, many marine phenomena, including ocean oil spills, appear as dark spots. In order to identify marine oil spill from many dark spots, the geometrical and statistical distribution features of dark spots need to be extracted in the SAR image to serve as the basis for further classification (identification) of marine oil spills. The application of the segmentation method based on geometric partition technology Feature extraction of dark spots in SAR images. First, the model of high-resolution SAR images is established, then the geometric parameters and statistical distributions of parameters are extracted by maximizing expectation and M-H algorithm. Experimental results show that this method not only can accurately extract the geometry of dark spots, but also can effectively estimate the statistical distribution parameters.