The Research On Image Coding And Decoding Technology Based On Wavelet Transform

As the primary carrier image play an important role in multimedia technology when people try to acquire the information about outside. However, huge information of image bring lots of difficulties at their storage and transmission, thus image compress have been an active research area. Wavelet coding technique as new generation image coding technique has good foreground of application. Wavelet transformation have local feature of time and frequency, multiresolution characteristic that can combine human vision system taking on good value of application.Firstly, the text introduces the meaning of image coding and developing process of wavelet coding, analyses correlation theory of wavelet transformation and bring multiresolution analysis into image coding that supply theoretical principle for wavelet image coding technique. Secondly, I have comprehended Embed Zerotree Wavelet algorithmic profundity by analyzing two zero tree coding arithmetic-EZW and SPIHT. Most important characteristic of Embed Zerotree Wavelet coding is that coding process can stop at any position according to request of target image. So it is suitable for progressive transmission of image and reduces transmission bandwidth effectively. After that, to point against insufficiency of EZW, I make some efforts to improve the algorithmic including two aspects, one shall code low-frequency band that concentrate most of energy with lossless coding technique, based on correlation theory of HVS, introducing concept of edge threshold, pay attention to extract the edge information of image and make reconstructed image clear in low code rate. Finally, I set up software flow chart of improvement algorithm, from two aspects, objective PSNR and subjective vision, inspect and verify algorithmic validity.