Data-Driven Progressive Mesh Compression Using Associated Properties

Many 3D models carry associated properties, such as normals or colors, with the geometry and connectivity data. Existing progressive mesh compression techniques usually do not take these properties into account during compression, missing the quality improvement of the intermediate decompression meshes. In this work, we propose a novel progressive compression algorithm that uses the associated properties of the mesh to drive the compression process. Any property can be used as long as a distance function is defined for that particular property. The algorithm builds a kd-tree structure using a voxelisation process, which recursively separates the set of vertices according to the associated properties distances, resulting in an improved quality of the intermediate meshes. We evaluate our method by comparing its compression ratios to recent algorithms, and by conducting a perceptive evaluation. At equal rates, our method provides a better overall visual quality.