Research On The Wavelet Transform Image Compression

The subject is from Project 60702049 supported by NSFC.The application of wavelet transform in image compression has attracted more and more researchers, since the issue of the new static image compression standard--JPEG 2000. Compared with JPEG standard, the wavelet based image compression method not only improves the compression rate by 30%, but also has the advantage of multi-resolution representation and scalability.Though the wavelet transform has enhanced the compression quality to a great extent, there are still many aspects to be researched and optimized. This paper aims to design a suitable wavelet packet transform in JPEG 2000, to develop the compression quality, based on the study of the wavelet transform and the wavelet image coding. To get the target, the paper studied the characteristic of the wavelet transform and the advantage of its compression application, compared three classic image compression algorithms, including EZW(Embedded Zerotree Wavelet), SPIHT(Set Partitioning in Hierarchical Trees) and EBCOT(Embedded Block Coding with Optimized Truncation).Based on the research above, this paper firstly designed an adaptive wavelet packet transform applied on JPEG 2000. Its cost function is calculated by estimating the coding length of EBCOT.Then the improved JPEG 2000 coder and decoder were built to test the wavelet packet transform. Experimental results of PSNR and coding time have been analyzed in detail.Lastly, to promote the wavelet packet transform compression quality, the wavelet filters have been researched and the wavelet coefficients have been optimized. The final experimental result shows that the optimized wavelet packet transform can improve the compression PSNR and the reconstruction quality.