Research On Image Compression Based On Wavelet And Fractal

Image is some important kind of information carrier in our life and one image usually contains more abundant information than other Medias, such as words. But the raw digital images are very big in size. If the image data couldn't be compressed efficiently, it will not meet the needs of high speed transmission and efficient storage nowadays. So, image compression, as a crucial, promising and vigorous technology in the fields of communication and multimedia, is rapidly expanding now.The coding technology of fractal compression and wavelet transform are new image compression for the latest one decade. People pay more attention to the excellence of link of both technologies. The article is deeply explore along the way, and did some researches. Firstly, two basic fractal coding algorithms are researched. Simulations show that Quad-trees algorithm is superior to Jacquin algorithm in coding complexity, but can not solve the high coding complexity. Secondly, Selection rules suits for image compression coding wavelet base is discussed, pointing out the wavelet base which has good linear phase and regularity and big vanishing movement is the first choice. Simulation results verified that the quality of the wavelet image compression is in direct relation with the selection of decomposing level and threshold. That is, under the same decomposition level, the bigger the threshold, the higher the compression rate. But when all the high-frequency components were discarded, with the threshold increasing the compression ratio achieves the limit no longer to increase. Finally, based on multi-resolution property of wavelet transform and some fractal characteristic of the image, combining fractal coding method, an image compression algorithm based on wavelet and fractal is proposed. In the algorithm, DCT coding is used in the low frequency part, and basic fractal compression coding is used in the high frequency part. Simulation indicates that, under the premise that reconstructed image quality is guaranteed, the coding time decreased greatly, and the compression rate is improved.