20th International Conference on 3D Web Technology (Web3D 2015) Paper

Federico Ponchio, Matteo Dellepiane – Visual Computing Lab CNR ISTI Pisa – “Fast decompression for web-based view-dependent 3D rendering.” Real time geometry processing has progressively reached a performance level that makes a number of signal-inspired primitives practical for on-line applications scenarios.  This often comes through the  joint  design  of  operators,  data  structure  and  even  dedicated hardware.   Among  the  major  classes  of  geometric  operators,  filtering and super-sampling (via tessellation) have been successfully expressed  under  high-performance  constraints.   The  subsampling operator i.e., adaptive simplification, remains however a challenging  case  for  non-trivial  input  models.   In  this  paper,  we  build  a fast geometry simplification algorithm over a new concept:  Morton Integrals.  By summing up quadric error metric matrices along Morton-ordered surface samples,  we can extract concurrently the nodes of an adaptive cut in the so-defined implicit hierarchy, and optimize  all  simplified  vertices  in  parallel.
This  approach  is  inspired by integral images and exploits recent advances in high performance spatial hierarchy construction and traversal.  As a result, our  GPU  implementation  can  downsample  a  mesh  made  of  sev eral millions of polygons at interactive rates, while providing better quality than uniform simplification and preserving important salient eatures. We present results for surface meshes, polygon soups and point clouds, and discuss variations of our approach to account for per-sample attributes and alternatives error metrics. Efficient  transmission  of  3D  data  to  Web  clients  and  mobile applications remains a challenge due to limited bandwidth. Most of the research focus in the context of mesh compression has been on improving compression ratio.  However, in this context the use of Javascript on the Web and low power CPUS in mobile applications led to critical computational costs.  Progressive decoding improves the user experience by providing a simplified version of the model that  refines  with  time,  and  it’s  able  to  mask  latency.    Current approaches do so at very poor compression rates or at additional computational  cost.    The  need  for  better  performing  algorithms is  especially  evident  with  this  class  of  methods  where  Limper [Limper et al. 2013b] demonstrated how decoding time becomes a  limiting  factor  even  at  moderately  low  bandwidths. In  this paper we present a novel multi-resolution WebGL based rendering algorithm  which  combines  progressive  loading,  view-dependent resolution and mesh compression, providing high frame rates and a decoding speed of million of triangles per second in Javascript. This method is parallelizable, robust to non-manifold meshes, and scalable to very large models. (PDF)