The Research And Development Based On Wavelet Zero-tree Embedded Image Coding Algorithm

The technique of wavelet transform has been applied and studied in the image coding field extensively because of its good feature of time-frequency and human visual system. Especially, aiming at the tree-structure of wavelet coefficients, many classical image coding algorithms appear in succession.But there are also lots of disadvantages for these methods, therefore the research on these methods is very significant.In the paper, firstly, the theories about image coding and wavelet transform are dissertated. And the embedded wavelet zero-tree image coding(EZW) and its improved algorithms are discussed and analysed in detail. Secondly, based on the research on the selection of wavelet basis, confirmation of decomposition stages and border processing, many important conclusions are concluded. Finally,a new image coding algorithm called "embedded image coding methods based on wavelet-tree energy" is offered and simulated by matlab emulation soft. The emulation results indicate the new algorithm can largerly improve the peak signal-to-noise ratio (PSNR) of the reconstructed image compared to EZW. The main research contributions and conclusions are summarized as follows.(1) obtain better reconstruct image choosing the bi-orthogonal (9,7) wavelet basis and symmetric periodic extension.(2) offer a conception "wavelet-tree energy" and use this conception to contrive scan-order of the improved image coding method.(3) contrive new methods about scan and virtual value mapping which can effectively eliminate the redundancy consisting in the same leave.