An enhanced density peak-based clustering approach for hyperspectral band selection
An enhanced density peak-based clustering approach for hyperspectral band selection
复制标题
DOI：
10.1109/igarss.2015.7325966
复制
发表时间：
2015-07
期刊：
影响因子：
--
通讯作者：
Guihua Tang;Sen Jia;Jun Yu Li
中科院分区：
文献类型：
--
作者：
Guihua Tang;Sen Jia;Jun Yu Li
最近，一个快速的密度峰值为基础的聚类算法，即FDPC，已经证明了它的非球形聚类问题的权力。在本文中，我们提出了一种增强的快速密度峰值聚类，即E-FDPC，超谱波段选择。与原FDPC相比，所提出的E-FDPC的主要贡献是两倍。首先，我们引入一个参数来控制归一化的局部密度和簇内距离之间的权重。另一方面是，我们提出了一个基于指数的学习规则，以调整不同数量的选定波段的截止阈值，它是经验定义的FDPC。此外，一个有效的策略，称为孤立点停止准则，自动确定适当的波段数。也就是说，聚类过程将因孤立点（一个聚类中的唯一点）的出现而停止。在真实的高光谱数据上的实验结果表明，E-FDPC方法比FDPC和其他波段选择方法具有更高的分类精度。
Recently, a fast density peak-based clustering algorithm, namely FDPC, has demonstrated its power on nonspherical clustering problems. In this paper, we propose an enhanced fast density peak-based clustering, namely E-FDPC, for hy-perspectral band selection. The main contributions of the proposed E-FDPC, in comparison with the original FDPC are two folds. First, we introduce a parameter to control the weight between the normalized local density and intra-cluster distance. The other aspect is that, we present an exponential-based learning rule to adjust the cut-off threshold for different number of selected bands, where it is empirically defined in FDPC. Furthermore, an effective strategy, called isolated-point-stopping criterion, is developed to automatically determine the appropriate number of bands. That is, the clustering process will be stopped by the emergence of the isolated point (the only point in one cluster). Experimental results on real hyperspectral data demonstrate that E-FDPC approach could achieve higher overall classification accuracies than FDPC and other state-of-the-art band selection techniques.