Listless Zerotree Image Coding Algorithm Based On SPIHT

Since the late 1980s, fast Wavelet transform is the introduction of image processing, because of the time and frequency domain good localized performance, easy-to-take into account the human visual system, as well as good to relative performance, and are widely used in image compression field, and has obtained good results, as the main image compression technology.Wavelet transform coding in general there are two categories:one category is tree-structure algorithm, a block-type structure algorithms. Among the tree-structure algorithms because of its simple and effective and compression ratio controllable, etc., laid it in image compression. This article is based on this type of encoding algorithm to expand research, after an overview of the basic theory of image compression encoding and wavelet transform, details the three trees-type structure image coding: embedded zerotree wavelet,set partitioning in hierarchical trees listless zerotree coding, and on each encoding algorithm analysis was made of the advantages and disadvantages。SPIHT Algorithm for fast Wavelet transformation process, you need to do a lot of convolution, computational complexity, real-time effect compression algorithm for DSP real-time implementation of increasing difficulty. Decoding algorithm requires three collections linked list stores the encoded information, memory consumption, not conducive to the hardware implementation. In the sorting and sequential transmission Wavelet coefficients, does not take full advantage of the wavelet transform energy aggregation, reduced quality image restoration。LZC algorithm demands less memory capacity than SPIHT algorithm. But thanks to the depth of the search strategy, reducing the performance of the encoder. Moreover, the algorithm uses a recursive structure implementation, thereby increasing the complexity of the algorithm, encoding speed down, as DSP real-time implementation increases degree of difficulty.For the deficiency of SPIHT and LZC algorithm, a new listless zerotree coding algorithms is proposed. The algorithm is on the basis of SPIHT, redefines the zerotree structure, while making full use of coefficient correlation of each sub-band after wavelet decomposition, uses the correlation between adjacent coefficients in the lowest frequency sub-band; inducts lifting Wavelet transform to construct biorthogonal D9/7 wavelet; improves the sequence of coding; and absorbs the idea of state bitmap in LZC algorithm.This article tested three standard gray-scale images (Lena, Goldhill and Barbara). useing the new algorithm and SPIHT algorithm. The results showed that:in the case of low bit rate, the edge distortion of the reconstructed images obtained by SPIHT algorithm was serious, some details had not resolved, but that of the reconstructed images obtained by the new algorithm was not significant, visual effect of the reconstructed images obtained by new algorithm was more excellent than that of the reconstructed images obtained by SPIHT algorithm; the PSNR vaule of the reconstructed images obtained by new algorithm was higher than that of the reconstructed images obtained by SPIHT algorithm.We can conclude from the above that:the overall performance of the new algorithm is better than SPIHT algorithm.