基于改进FCM的极化SAR图像机场跑道区域检测

Airport Runway Area Detection for SAR Image Based on Improved FCM

该文提出了一种基于改进FCM的极化SAR图像机场跑道区域检测方法。针对传统FCM算法对噪声敏感、未考虑相邻像素间的关联的问题,该文引入Wishart距离替代欧式距离,并用邻域像素信息对原始隶属度函数进行加权,提高了算法的分类性能。接着根据跑道区域的弱散射特性从模糊聚类结果中提取感兴趣区域。最后利用跑道的结构特征对感兴趣区域进行辨识。实验结果表明,该文算法能够有效检测出机场跑道区域,结构较完整。

An improved FCM-based airport runway area detection method for polarimetric SAR image is proposed in this paper.Aiming at the problem that the traditional FCM algorithm is sensitive to noise and does not consider the correlation between adjacent pixels,Wishart distance is introduced to replace euclidean distance,and the original membership function is weighted with neighborhood pixel information,both of which can improve the classification performance of the algorithm.Then the regions of interest are extracted from the fuzzy clustering results according to the weak scattering characteristics of the runway region.Finally,the region of interest is identified by utilizing the structural features of runway.The experiment results show that the algorithm can effectively detect the airport runway area with complete structure.