Rate-Distortion Optimized Wavelet-based Irregular Mesh Coding

Abstract

Abstract:This work investigates the optimization of mesh quality at lossy rates for a lossless scalable wavelet-based irregular mesh codec. Whereas previously proposed wavelet-based irregular mesh codecs offer coarse-grain resolution scalability, in this paper we propose a coding scheme which enables fine-grain quality scalability. This is done by avoiding the use of geometric data in the encoding process, which reduces dependencies within the data stream and allows for an unrestricted storage and transmission order of wavelet subband bitplanes and connectivity information. This in turn allows us to perform rate-distortion optimization, whereby the subband bitplanes to be encoded are determined by minimizing distortion subject to an overall target bitrate. Experimental results show that the proposed coding approach offers fine-grain quality scalability, achieves optimality in rate-distortion sense and improves compression performance over the state of the art.

Scalable Wavelet-Based Coding of Irregular Meshes With Interactive Region-of-Interest Support

“…For irregular meshes, each wavelet subband W j−1 (see Figure 2) contains the geometry information represented by the conventional wavelet coefficients w g ∈ G j−1 obtained via lifting, together with explicit connectivity information w c ∈ C j−1 to ensure the correct connectivity. For encoding, we employed a so-called template mesh, proposed originally in our previous works [7], [35], which was introduced in order to decouple the transform step from the encoding step and to allow for quality scalability. For each subband j − 1, a template mesh M T j−1 is maintained, representing all connectivity information in the original mesh.…”

“…• successive approximation quantization (SAQ) of the wavelet coefficients [36] and bitplane coding, performing scalable quantization and coding of the wavelet coefficients and enabling quality and resolution scalability [7], [35] over entire models, without providing ROI support; • wavelet coefficient boosting [10], discussed in Section IV, enabling encoder-side ROI support; • the adaptive wavelet transform proposed in Section V, which is a key component to enable decoder-side ROI coding support; • dynamic tile-based coding proposed in Section VI, enabling interactive ROI-support; and • rate-distortion optimization, detailed in Section VII, offering optimized allocation of rate across wavelet subbands and bitplanes.…”

“…In [7], [35], we used the template mesh for embedding both the connectivity and geometry samples in R 3 as described above. In Section V, the same template mesh is additionally used to map the wavelet coefficient supports S(w) from M T j where they are ensured to be accurately represented, to M α≤j where these supports are not necessarily fully reconstructed.…”

“…With this tiling available, we propose a rate-distortion optimization of the coding system which aims at optimally allocating rate across different tiles. We have investigated rate-distortion optimization for untiled wavelet-based irregular mesh encoding in [35]. In this paper, we generalize this approach for dynamic tiling, as detailed next.…”