Improving image compression performance with balanced multiwavelets

The multiwavelet transform - unlike the scalar wavelet transform - allows orthogonality and symmetry to coexist. For lossy image compression, the balancing order of the multiwavelet filter bank dictates energy compaction. But balancing alone does not guarantee good compression performance. Filter bank characteristics such as shift-variance and magnitude response also influence peak signal-to-noise ratio (PSNR) and perceived image quality. In this paper we analyze the effect of these multiwavelet properties on image compression. Our results indicate the best multiwavelet image compression performance reported to date.