[Zurück]

S. Biasotti, A. Cerri, D. Giorgi:
"Robustness and modularity of 2-dimensional size functions";
Vortrag: CAIP 2011 - 14th International Conference on Computer Analysis of Images and Patterns, Sevill, Spanien; 29.08.2011 - 31.08.2011; in: "CAIP 2011 - 14th International Conference on Computer Analysis of Images and Patterns", Proc. CAIP 2011 - 14th International Conference on Computer Analysis of Images and Patterns, LNCS 6854, Ainhoa Berciano et al. Eds./Springer, Seville, Spanien (2011), ISBN: 3642236715; S. 34 - 41.

This paper deals with the concepts of 2-dimensional size function and 2-dimensional matching distance. These are two ingredients of (2-dimensional) Size Theory, a geometrical/topological approach to shape analysis and comparison. 2-dimensional size functions are shape descriptors providing a signature of the shapes under study, while the
2-dimensional distance is the tool to compare them. The aim of the present paper is to validate, through some experiments on 3D-models, a computational framework recently introduced to deal with 2-dimensional Size Theory. We will show that the cited framework is modular and robust with respect to noise, non-rigid and non-metric-preserving shape transformations. The proposed framework allows us to improve the ability of 2-dimensional size functions in discriminating between shapes.