In this paper the Hausdorff distance, and a robust modified variant of the Hausdorff distance are used for the purpose of matching graphs whose structure can be described in terms of triangular faces. A geometric quantity from the geodesic triangle and the corresponding Euclidean triangle is deduced and used as a feature for the purposes of gauging the similarity of graphs, and hence clustering them, we experiment on sets of graphs representing the proximity image features in different views of different objects from the CMU, MOVI and chalet house sequences. By applying multidimensional scaling to the Hausdorff distances between the different object views, we demonstrate that this representation is capable of clustering the different views of the same object together.
展开▼
