In this paper we exploit the effectiveness of bioinformatics tools to deal with 3D shape matching. The key idea is to transform the shape into a biological sequence and take advantage of bioinformatics tools for sequence alignment to improve shape matching.In order to extract a reliable ordering of mesh vertices we employ the spectral-based sequencing method derived from the well known Fiedler Vector. Local geometric features are then collected and quantized into a finite set of discrete values in analogy with nucleotide or aminoacid sequence. Two standard biological sequence matching strategies are employed aiming at evaluating both local and global alignment methods.Preliminary experiments are performed on standard non-rigid shape datasets by showing promising results in comparison with other methods.
Bicego, M., Danese, S., Melzi, S., Castelani, U. (2015). A bioinformatics approach to 3D shape matching. In Proc. ECCV NORDIA 14 (pp.313-325). HEIDELBERGER PLATZ 3, D-14197 BERLIN, GERMANY : Springer Verlag [10.1007/978-3-319-16220-1_22].
A bioinformatics approach to 3D shape matching
Melzi, S;
2015
Abstract
In this paper we exploit the effectiveness of bioinformatics tools to deal with 3D shape matching. The key idea is to transform the shape into a biological sequence and take advantage of bioinformatics tools for sequence alignment to improve shape matching.In order to extract a reliable ordering of mesh vertices we employ the spectral-based sequencing method derived from the well known Fiedler Vector. Local geometric features are then collected and quantized into a finite set of discrete values in analogy with nucleotide or aminoacid sequence. Two standard biological sequence matching strategies are employed aiming at evaluating both local and global alignment methods.Preliminary experiments are performed on standard non-rigid shape datasets by showing promising results in comparison with other methods.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.