Compression and Transmission of Multi-Resolution Clustered Meshes

When working with large geometric models in a client-server environment, it is desirable to generate or store hierarchies of levels of detail (LOD) on the server and to progressively transmit them to the client in a compressed form. In addition to the traditional role of reducing polygon counts to accelerate client frame rates, such an approach permits the LOD hierarchy to be used to reduce the transmission latency. Among the many existing methods for generating multi-resolution polygonal models only a few address progressive transmission, efficient compression, or a combination of the two. With the exception of the Progressive Simplicial Complex (PSC) method, all of the existing compression schemes are limited to manifold meshes with constant topology and all of the existing schemes which handle non-manifold meshes feature little or no compression. In this paper, we introduce a new scheme for compressing and progressively transmitting (potentially non-manifold) LODs as Multi-Resolution Clustered (MRC) meshes. We extend the Topological Surgery bit-stream syntax to handle MRC meshes. Connectivity information from low to high resolution.. The method is progressive with reasonable transmission latencies, produces compression ratios comparable with the most efficient single-resolution schemes, and is significantly more efficient then the PSC method.