The Research Of Image Compression Based On Wavelet Zerotree And Fractal

With the development of information network and multimedia technology, digital image, as a kind of the most important media, is used much widely. Image compression becomes more and more important for great deal of data. In this paper, some most popular compression coding: Wavelet Zerotree coding and Fractal coding and are introduced, the advantages, disadvantages and developments of these algorithms are discussed. The improvements are made on SPIHT and the fractal image coding in DCT domain. Finally, the conclusions and expectations of image compression algorithms are given. The main work includes the following four aspects:On the research of SPIHT, this paper presents an improved SPIHT algorithm on the basis of 9-7-tap integer wavelet transform. In the improved algorithm, at first the image is decomposed by using integer wavelet, then the low-frequency sub-image wavelet coefficients are processed using predictive coding, meanwhile, the high-frequency ones are proposed changing scanning mode to obtain the biggest coefficient and using frequency- priority to export the coefficients. The result shows for the same output bit rate, this algorithm not only makes the image compressed effectively with high PSNR, but also in an efficient way to reduce processing time .On the research of the fractal image coding in DCT domain, this paper presents an improved algorithm. In the improved algorithm, at first, plain blocks are selected based on human visual system. Then by selecting out the uniform parts in the DCT domain, the direct current encode directly without any fractal matching. For matching time reducing, the mean squared error computations of the eight orientations of the domain blocks are simply reduced into two groups of inner products in the DCT domain. The result shows for the same matching errors, this algorithm not only makes the image compressed effectively with high PSNR and much better visual effect, but also in an efficient way to reduce processing time.Based on the characteristic structure of wavelet coefficients, combined with two characters, we use fractal and zerotree coding to code the image. When we use the fractal coding, we change the conventional MSE because there is no direct current in the wavelet coefficients and we use the error emendation matrix to revise wavelet coefficients. Theoretical analysis and experimental results presented in the paper show that there are higher compressing rate and quality of the image than the commonly fractal and zerotree coding methods.