Using k-MST, k-EC and k-VC neighbor graphs constructionmethods with spatial coherent distance for manifold learningin hyperspectral image processing

摘要

Dimensionality reduction is a key step in hyperspectral image processing. Recent investigations are looking intononlinear manifold learning for dimensionality reduction in hyperspectral imagery. Nonlinear manifold learningmethods such as Isomap, and Local Linear Embedding are use to recover the low dimensional representation of anunknown nonlinear manifold for high dimensional data where it is important to retain the neighborhood structureof the manifold. Although these algorithms use different philosophies for recovering the nonlinear manifold, theyall incorporate neighborhood information from each data point to construct a weighted graph having the datapoints as vertices. Thus the performance of these methods highly depends on how well these neighborhoods areselected, since all subsequent steps rely on it. The k-NN algorithm is the most widely used technique for neighborselection in manifold learning. However, it can result in a disconnected graph and it does not fully exploit spatialneighborhood information from the image in selecting points to form tha neighborhoods. In this paper, recentlyproposed methods for constructing the weighted graph in manifold learning are studied: k-VC, k-EC and k-MST, which have the advantage of creating connected graphs and have performed well in artificial data sets.Spatial information of the hyperspectral images is included in the manifold learning process by using spatialcoherence. Experiments are conducted with artificial data and hyperspectral images. For hyperspectral images,classification accuracy was used as an indirect measure to understand how the low dimensional embedding ofthe data reduces dimensionality while still maintaining good discrimination performance.