A Dynamical Systems Algorithm for Clustering in Hyperspectral Imagery
This provides a new tool for understanding hyperspectral data in high dimensions, though it appears incremental as it builds on existing clustering concepts.
The paper tackles clustering in hyperspectral imagery by proposing a dynamical systems algorithm that pushes data points toward dense regions to form clusters, automating the number of classes and achieving high accuracy, with evaluation on the Urban scene showing competitive performance against k-means using ground truth materials.
In this paper we present a new dynamical systems algorithm for clustering in hyperspectral images. The main idea of the algorithm is that data points are pushed\' in the direction of increasing density and groups of pixels that end up in the same dense regions belong to the same class. This is essentially a numerical solution of the differential equation defined by the gradient of the density of data points on the data manifold. The number of classes is automated and the resulting clustering can be extremely accurate. In addition to providing a accurate clustering, this algorithm presents a new tool for understanding hyperspectral data in high dimensions. We evaluate the algorithm on the Urban (Available at www.tec.ary.mil/Hypercube/) scene comparing performance against the k-means algorithm using pre-identified classes of materials as ground truth.