3D Volume Data Compression Based on Adaptive Wavelet

Volume data compression is very important for transmission and display of very large 3D volume data sets. Based on adaptive wavelet, an efficient compression scheme was proposed, which mainly exploited intra-band dependencies of wavelet coefficients. Blockwise hierarchy wavelet decomposition was performed to obtain approximate and detail coefficient block. Because significant coefficients were clustered in certain areas in wavelet domain, corresponding to the edge and textual regions in the volume, we repeatedly applied a process in which each coefficient block was subdivided into eight sub-blocks if there were significant coefficients in the block. Finally, we used one bit to indicate block or sub-block significant map, resulting in fewer bits for storing significant map and that all the coefficients stored in data area were significant. Due to adopt adaptive wavelet in which homogeneous regions are smoothed while discontinuities are preserved, high quality image can be obtained at low resolution, which is useful for transmission between data server and browsing clients over low-bandwidth networks