8 February 2012 Reeb graph computation through spectral clustering
Author Affiliations +
Abstract
The Reeb graph provides a structure that encodes the topology of a shape, and it has been gaining in popularity in shape analysis and understanding. We introduce a spectral clustering method to compute the Reeb graph. Given a 3-D model embedded in the Euclidean space, we define the Morse function according to the connected components of the 3-D model in a spectral space. The spectral clustering formulation gives rise to a consistent Reeb graph over pose changes of the same object with meaningful subparts and yields a hierarchical computation of the Reeb graph. We prove that this method is theoretically reasonable, and experimental results show its efficiency.
© 2012 Society of Photo-Optical Instrumentation Engineers (SPIE)
Teng Ma, Teng Ma, Zhuangzhi Wu, Zhuangzhi Wu, Pei Luo, Pei Luo, Lu Feng, Lu Feng, } "Reeb graph computation through spectral clustering," Optical Engineering 51(1), 017209 (8 February 2012). https://doi.org/10.1117/1.OE.51.1.017209 . Submission:
JOURNAL ARTICLE
10 PAGES


SHARE
Back to Top