The same-level subbands correlation removed based SPIHT image coding algorithm

A new SPIHT (set partitioning in hierarchical trees) image compression-coding algorithm based on the same-level subbands correlation removed (SCR-SPIHT) is presented. The SCR-SPIHT disposes the lowest-frequency subband and the high-frequency subbands respectively. Firstly, code the lowest-frequency subband using DPCM (Differential Pulse Code Modulation). Secondly, define the conception of Correlation Node according to the correlation of the nonsignificant coefficients in the same-Level high-frequency subbands and the same high-frequency subband, and the extended space orientation tree based on the Correlation Node is constructed. Here, the Correlation Node is composed of four adjacent pixels of 2×2 in the same subbands and eight adjacent pixels of 2×2×2 in the same location of the same-Level subbands. The symbols of EZW (Embedded Zero-tree Wavelet) and the new symbol based on EZW were applied to the extended space orientation tree, and the three lists of SPIHT (LIS, LIP, and LSP) are substituted with two lists of pixel states, which saved the memory of program. Simulation experiments showed that this algorithm not only saved more coding time, but also had better performance than SPIHT algorithm.