Discrete minimum distortion correspondence problems for non-rigid shape matching

Chaohui Wang (1, 2) , Michael M. Bronstein (3) , Nikolaos Paragios (1, 2)

1 MAS - Mathématiques Appliquées aux Systèmes - EA 4037
2 GALEN - Organ Modeling through Extraction, Representation and Understanding of Medical Image Content
3 Department of Computer Science [Haifa]

Similarity and correspondence are two fundamental archetype problems in shape analysis, encountered in numerous application in computer vision and pattern recognition. Many methods for shape similarity and correspondence boil down to the minimum-distortion correspondence problem, in which two shapes are endowed with certain structure, and one attempts to find the matching with smallest structure distortion between them. Defining structures invariant to some class of shape transformations results in an invariant minimum-distortion correspondence or similarity. In this paper, we model shapes using local and global structures and formulate the invariant correspondence problem as binary graph labeling. We perform challenging non-rigid shape matching experiments, and show how different choice of structure results in invariance under various classes of deformations.

Mots clés: shape recognition; shape similarity; Gromov-Hausdorff metric

Domaines: Vision par ordinateur et reconnaissance de formes [cs.CV]

Chaohui Wang, Michael M. Bronstein, Nikolaos Paragios. Discrete minimum distortion correspondence problems for non-rigid shape matching. [Research Report] RR-7333, INRIA. 2010. ⟨inria-00498591⟩