Locally optimal codebook design for quadtree-based vector quantization
Locally optimal codebook design for quadtree-based vector quantization
复制标题
基于四叉树的矢量量化的局部最优码本设计
DOI：
10.1109/icassp.1995.480051
复制
发表时间：
1995
期刊：
影响因子：
--
通讯作者：
S. Mitra
中科院分区：
文献类型：
--
作者：
M. Lightstone;K. Rose;S. Mitra
讨论了基于四叉树的矢量量化器的优化设计。到目前为止，该领域的工作主要集中在优化给定叶子量化器的四叉树结构上，很少关注量化器本身的设计。在考虑叶量化器的情况下，对码本进行优化，而不考虑最终的四叉树分割。然而，单独考虑每个问题是不够的，因为单独的优化会导致整体的次优解决方案。相反，四叉树结构和叶码本的联合设计必须考虑整体的最优性。我们建议的方法是一种“四叉树”约束版本的熵约束向量量化设计方法。为此，推导了叶码本的质心条件，该条件代表了可变速率四叉树编码的必要最优性条件。这种情况下，当迭代Sullivan和Baker的最优四叉树分割策略(参见IEEE Trans。在图像处理，vol.3, p.327-331, 1994年5月)的结果单调下降的率失真代价函数，因此，一个（至少局部）最优的四叉树解决方案。
The optimal design of quadtree-based vector quantizers is addressed. Until now, work in this area has focused on optimizing the quadtree structure for a given set of leaf quantizers with little attention spent on the design of the quantizers themselves. In cases where the leaf quantizers were considered, codebooks were optimized without regard to the ultimate quadtree segmentation. However, it is not sufficient to consider each problem independently, as separate optimization leads to an overall suboptimal solution. Rather, joint design of the quadtree structure and the leaf codebooks must be considered for overall optimality. The method we suggest is a "quadtree" constrained version of the entropy-constrained vector quantization design method. To this end, a centroid condition for the leaf codebooks is derived that represents a necessary optimality condition for variable-rate quadtree coding. This condition, when iterated with the optimal quadtree segmentation strategy of Sullivan and Baker (see IEEE Trans. on Image Processing, vol.3, p.327-331, May 1994) results in a monotonically descending rate-distortion cost function, and consequently, an (at least locally) optimal quadtree solution.