3D TRIANGLE MESH COMPRESSION BASED ON VECTOR QUANTIZATION WITH k-RING VECTOR PREDICTION

The transmission and storage of large amounts of triangle and vertex geometry data are required for rendering geometrically detailed 3D meshes. To alleviate bandwidth requirements, this paper uses vector quantization (VQ) as an effective lossy vertex data compression technique for triangle meshes with high rate-distortion performance. The proposed novel VQ-based vertex encoding algorithm adopts a region growing based k-ring prediction scheme. During the encoding process, the value k related to the current vertex to be encoded is estimated by the prediction errors of the preceding encoded vertices, and then the adaptive prediction rule for the current vertex is obtained according to the value k. Therefore, the dynamic range of the residual vectors generated by the proposed prediction scheme is reduced. Thus it is expected to achieve a higher quantization quality with the same codebook size. Experimental results show that, compared with the vertex encoding algorithm based on the conventional parallelogram prediction scheme, the proposed algorithm can achieve a higher encoding quality at the same bit rate.