Vector Set Partitioning with Classified Successive Refinement Vq for Embedded Wavelet Image Coding*

Set Partitioning in Hierarchical Trees (SPIHT), proposed by Said and Pearlman [1], is generally regarded as a very efficient wavelet-based still image compression scheme. The algorithm uses an efficient, joint scanning and bit-allocation mechanism for quantizing the scalar wavelet coefficients, and produces a perfectly embedded bitstream. This work extends set partitioning to scan vectors of wavelet coefficients, and use successive refinement VQ techniques such as multistage and tree-structured VQ, to quantize several wavelet coefficients at once. The new scheme is named VSPIHT (Vector SPIHT). Coding results are presented to demonstrate that the vector-based approach (without arithmetic coding) surpasses the scalar counterpart (also without arithmetic coding), in the mean-squared-error sense, for most images at low bitrates. The superiority of the vector-based approach is more pronounced for images that are generally regarded as difficult to code (such as Barbara) because of a large amount of detail.