摘要
针对传统近邻传播(Affinity Propagation,AP)聚类算法使用欧式距离构建相似度矩阵,不能有效描述极化SAR数据复杂分布的问题,本文提出一种新的基于联合流形距离的AP聚类算法(CMD-AP)用于极化SAR图像分类。首先将待分类极化SAR图像分割成若干超像素,在相应的极化特征基础上加入图像纹理特征,利用拉普拉斯特征映射算法对特征降维;然后结合相干矩阵Wishart流形和特征矢量欧式流形作为流形距离测度,构造相似性矩阵;最后利用上述相似性矩阵,采用AP聚类算法,对极化SAR图像进行分类。该算法充分考虑了极化SAR数据集潜在的流形结构,将联合的流形距离测度引入AP算法中。实验表明,本文算法提高了极化SAR图像的分类精度,具有更优的区域一致性和边缘保持效果。
A novel affinity propagation（ AP） clustering method based on combined manifold distance（ CMD-AP） is proposed for polarimetric SAR（ Pol SAR） image classification. It solves the problem that traditional AP clustering algorithm,which employs Euclidean distance as similarity measures,cannot effectively process the complex Pol SAR data. After dividing the Pol SAR image into superpixels,texture features as well as corresponding polarimetric features are extracted from Pol SAR image,and Laplacian eigenmap（ LE） is used to reduce the dimension of the features. Then Wishart manifold of coherency matrix and Euclidean manifold of feature vector are combined to form the manifold distance measures. Finally,the Pol SAR image is classified by AP clustering algorithm. The proposed method considers the potential manifold structure of PolSAR data and introduces the combined manifold distance measures into AP clustering. Experimental results demonstrate that the proposed method improves the classification performance,has advantages in region homogeneity and edge preservation.
出处
《信号处理》
CSCD
北大核心
2016年第2期135-141,共7页
Journal of Signal Processing
基金
国家重点基础研究发展计划(2013CB329402)
国家自然科学基金(61271302,61272282)
高等学校学科创新引智计划(111计划)(B07048)
关键词
图像分类
极化合成孔径雷达
流形学习
近邻传播聚类
超像素
image classification
polarimetric synthetic aperture radar
manifold learning
affinity propagation clustering
superpixel
