Balanced Multiwavelet and Its Application in Image Compressing

Orthogonality and symmetry are desirable transform properties in image compression. And most multiwavelet can hold the two properties concomitantly. While the choice of pre-filter is of the utmost importance and pre-filtering brings complexity of computation in applications. Thereby, balanced multiwavelet which can avoid pre-filtering is considered widely by researchers. A new approach on the construction of orthogonal balanced multiwavelet is proposed. It holds the properties of 1st approximation order, orthogonality, symmetry-antisymmetry, balance at the same time and was applied in image compression. The experimental result shows that a better effect was obtained using the balanced multiwavelet than GHM and CL multiwavelets with no profiltering and lower computational complexity. Especially it is useful for compressing of the "rubber" model of iris image