PT Journal
AU Amjoun, R
Strasser, W
TI Predictive-DCT Coding for 3D Mesh Sequences Compression
SO JVRB - Journal of Virtual Reality and Broadcasting
PY 2008
VL 5(2008)
IS 6
DI 10.20385/1860-2037/5.2008.6
DE Animation; DCT; animated mesh compression; clustering; local coordinate frame; predictive coding
AB This paper proposes a new compression algorithm fordynamic 3d meshes. In such a sequence of meshes,neighboring vertices have a strong tendency to behavesimilarly and the degree of dependencies between theirlocations in two successive frames is very large which canbe efficiently exploited using a combination of Predictiveand DCT coders (PDCT). Our strategy gathers meshvertices of similar motions into clusters, establish alocal coordinate frame (LCF) for each cluster andencodes frame by frame and each cluster separately. Thevertices of each cluster have small variation over atime relative to the LCF. Therefore, the location ofeach new vertex is well predicted from its location inthe previous frame relative to the LCF of its cluster.The difference between the original and the predictedlocal coordinates are then transformed into frequencydomain using DCT. The resulting DCT coefficients arequantized and compressed with entropy coding. Theoriginal sequence of meshes can be reconstructed from only a few non-zero DCT coefficients without significant loss in visual quality. Experimental results show that our strategy outperforms or comes close to other coders.
ER