Linear adaptive global node-tree filters based SPIHT MR image codec

During progressive transmission compressed images should be preserved for its significant features, different compression models available recently are unable to implement this. In propose paper, a novel method for low bit rate magnetic resonance image compression, implemented in three stages. In first stage, integer zero wavelet having unity positive and zero negative values are assigned, in second stage, multiple significant features are extended to zero integer wavelet decomposition for high bit rate values and positive-negative integer wavelet decomposition for low bit rate values. In third stage, during encoding bit stream header is initiated with absolute mean of sorting and refinement pass significant values. Multiple integer values are derived from SPIHT compression model. Proposed method of magnetic resonance image SPIHT codec had shown better results comparatively than SPIHT.