大幅面星载SAR图像中机场检测算法

重点研究了大幅面星载合成孔径雷达图像中的目标检测问题.选取机场区域作为检测目标,提出了一种基于区域形状特征的检测算法.算法使用自适应聚类分割法解决大幅面图像中复杂背景下的小目标分割问题,选取椭圆近似法代替常用的最小外接矩形和边界框法计算区域的尺寸和体态特征.通过对多幅实际获取自不同场景的大幅面星载合成孔径雷达图像进行实验,结果表明,本算法可快速、准确地检测出包含在场景中的单个或多个结构不同的机场区域.

Airport detection algorithm in large area satellite borne SAR images

This paper mainly discusses airport detection in the large area satellite-borne synthetic aperture radar(SAR) images.A new algorithm based on features of the regional shape is proposed.It employs an adaptive clustering method to solve the problem of small target segmentation from large area and complex background images.The ellipse-fitting method is utilized to approximate the scale and the shape of the considered region instead of the widely used approaches of the minimum enclosing rectangle and the bounding box.Several large area satellite-borne SAR images collected from separated scenes are used for experiment.The results show that the proposed algorithm is able to detect single or multiple airports with variable runways in a short time.