3D shape retrieval is becoming an acute issue for numerous applications that span from CAD to serious games to biomedicine and all contexts where it is fundamental to automatically retrieve geometric information from a collection of 3D models. This paper addresses 3D shape retrieval in terms of a graph-based description and the definition of a corresponding similarity measure. For this purpose, 3D models are represented as bags of shortest paths defined over well chosen Extended Reeb Graphs, while the similarity between pairs of Extended Reeb Graphs is addressed through kernels adapted to these descriptions. Results are comparable with the best results of the literature, and the modularity and evolutivity of the method ensure its applicability to other problems, from partial shape matching to classification.
More detail can easily be written here using Markdown and $\rm \LaTeX$ math code.