摘要
有限混合模型FM的分级聚类已广泛应用于不同领域,然而,由于它的计算复杂度与观测数据量平方成正比,致使在遥感影像方面应用受到了限制。另外,多光谱图像能提供空间和光谱两类信息详细的数据,但是,大多数多光谱图像聚类方法是基于像素的聚类,仅使用了其光谱信息而忽视了空间信息。本文定义一个相对混合密度函数,通过引入一个q-参数来调节各成分密度对其混合分布的贡献,提出一种广义有限混合模型GFM,设计一种新的适用于多光谱遥感影像的GFM分级聚类算法。该算法把MRF随机场和GFM模型结合在了一起,分类数通过PLIC准则自动确定。最后,利用仿真结果验证该算法的有效性,同时通过与K均值聚类、FM分级聚类以及SVMM分级聚类的比较说明本文算法的优越性。
Hierarchical clustering based on the Finite Mixture（FM） model has shown very good performance in a number of fields. However, it generally requires storage and computing at least proportional to the square of the dimension of observations, so that its application to large datasets has been hindered by a time and memory complexity. Otherwise, multispectral images provide detailed data with information in both the spatial and spectral domains. But many clustering methods for multispectral images are based on a per-pixel classification, while uses only spectral information and ignores spatial information. Firstly, a new mixture density function called the relative density function is defined. To adjust the contribution of the each component density to mixture density function, the q-parameter into the mixture densities is introduced. T he Generalized Finite Mixture （GF M） model is proposed in this paper. Also, a new hierarchical clustering based on GFM models, suitable for large datasets, e.g., multispectral remote sensing images, is proposed. This algorithm is integrated with GFM model and Markov random field. The number of clusters is automatically identified by using the Pseudolikelihood Information Criterion （PLIC）. At last, gives the simulation results, which testifies the validity of this algorithm. The experiment shows also a superior performance compared to several other methods, such as K-means and classical hierarchical clustering based on the classical FM model.
出处
《测绘学报》
EI
CSCD
北大核心
2007年第4期400-405,442,共7页
Acta Geodaetica et Cartographica Sinica
基金
国家"973"重点基础研究发展规划项目(2003CB716101)
