Research Of Wavelet-based Fractal Image Compression Arithmetic

Nowadays, much attention has been paid to two research branches in the field of image compression, that is, fractal coding and wavelet transformation coding. In particular, the hybrid algorithms of fractal and wavelet transformation coding have already been developed and are becoming popular. In this paper, we make some improvements on the application to image compression of the three research branches.It is well known that the encoding time to search matching domain block for every range block is definitely long and encoding speed is accordingly rather slow. In order to overcome such a problem, we propose and implement a new algorithm. In the process of matching domain blocks and range blocks, range block may encode with a smaller pieces, without considering its impact on compression rates, thus increasing the coding efficiency.We put forward a zerotree wavelet image compression algorithm with rigid package length which based on EZW(Embedded Zerotree Wavelet encoding algorithm). At first ,the datum of the image are turned into many zerotrees which consist of wavelet coefficients by using zerotree wavelet image compression algorithm. Then, those zerotrees are used to coefficient scanning, if meet the zerotree, all its' filial generation are omitted from zero. Which describe important factor location information greatly reduced, thus greatly improving the efficiency of the scanning code, enhanced coding efficiency.As is well know, in the general combining algorithm of fractal and wavelet transformation applying to image compression, one could product many subbands including one low frequency subband and several high frequency subbands after transforming the image by two-dimensional wavelet transformation one or more times. The similar high frequency subbands are assigned to encode by fractal coding algorithm. This algorithm has one advantage with high compression ratio, whereas it has one shortcoming with relatively slow encoding speed. In this paper, we propose the new algorithm of fractal and wavelet transformation, which is showed that it can improve the encoding speed by experimental results.In the end of this paper, we discussed the algorithms in the digital watermarks' application. Digital watermark technology is a wider application of information technology security, watermarks and the embedded information have a impact on the information security. This paper provides an approach based on a combination of fractal and wavelet about the digital watermarks technology. Experimental results...