ZERO TREE BASED ALGORITHM FOR STILL IMAGE CODING
Tian Guo · Chinese Journal of Computers · 1999
An improved zero tree based compression coding algorithm for still image is proposed in this paper. It can be viewed as an improved version of Shaprio's EZW algorithm. The coding stream includes three parts: (1) coding stream of the lowest frequency subband image, the lowest frequency subband concentrates most energy of original image, and is a representation of the information at all coarser scale. Therefore, small distortion for this subband may exert a tremendous influence to the original image. This subband is coded separately from other highpass subbands. The method applied to it is lossless or less loss coding algorithm so that to preserve all its information, and to get rid of the residue correlation; (2) the zero block pruning map:contrary to the EZW algorithm, the authors' scanning of the coefficients is performed in such a way that no parent node is scanned before its children. The significant map is created by pruning the zero trees from the spatial coefficient trees bottom to up; (3) the quantized data stream corresponding to the survivor nodes: use one proper scalar to quantify the significant coefficients uniformly, which will decrease not only computing complexity but also the bit costing for those survivor nodes coefficients. Several novel features of this scheme are achieved. Simulation results show that it is superior to the EZW algorithm in PSNR, visual effect and time complexity, etc.