摘要
提出了一种基于几何活动轮廓模型分割的SAR机场掩体目标检测方法。根据SAR图像的特性选择了Chan-Vese模型(C-V模型),对其进行加窗,使全局信息的C-V模型局部化,并扩展到多相,从而有效地分割出机场区域;再根据掩体目标的断头路位置特征检测目标潜在位置;最后利用掩体在SAR图像的强回波特性去除虚警,实现对掩体目标的有效检测。实验表明:该方法能克服相干斑的干扰,简化对复杂拓扑关系的处理,为检测机场掩体目标提供了一种有效技术途径。
A SAR airport shelter targets detection method based on the geometric active contour segmentation model has been put forward in this paper.The Chan-Vese model(C-V model) is chosen according to the characteristics of SAR image; a window is added to the C-V model to have local information and the model is extended to multiphase so that the airport area can be segmented effectively.Then the potential location of targets is detected according to the location of bunker targets.The dead end highway characteristic of bunker targets is just the location characteristic.At last,the echo of the bunkers of SAR image is used to remove false alarm and realize effective detection of the target bunkers.The experiment result show that this method can overcome the interference of coherent spot,simplify the processing of complicated topological relationship,and provide an effective technical way to detect the airport shelter targets.
出处
《测绘科学技术学报》
CSCD
北大核心
2013年第5期494-499,共6页
Journal of Geomatics Science and Technology
基金
军队探索项目(7131145)
关键词
图像分割
C-V模型
局部化
多相水平集
峰值特征
image segmentation
Chan-Vese model
localization
multiphase level set
peak power