Adaptive subband image coding considering spatial relationship on hierarchical pyramid
Tsuyoshi Otake, Akira Kawanaka · 2003
It is well known that the band-splitting method strongly influences the efficiency of subband image coding. Further, the coding efficiency is very dependent on the entropy-coding model. Particularly in the embedded zero tree wavelet (EZW), set partitioning in hierarchical trees (SPIHT), or the space-frequency quantization (SFQ) algorithm, the spatial dependency defined by the quad tree set of coefficients is employed and they give a good performance. In this work, we describe the adaptive decomposition method based on a rate-distortion characteristic of the quad tree set of coefficients that is independent of the band-splitting. Experimental results for some images show that our coder gives a better coding efficiency and it generally tends to be effective for octave decomposition in the lower bit rate application and for uniform decomposition in the high frequency subband in the higher case.