Research On Still Image Compression Based On Wavelet Transform

With the development of information and network technology, images and other media information, recorded and described, storage and transmission are all moving in the digital direction, and efficient image compression and transmission is receiving increasing attention. Therefore, the image compression technology become to the internationally popular research topic.The wavelet transform have a better partial and space-frequency charactristics, an ability of describing the non-stationary image signal and a good function of adapting to the eyes sense of visual characteristic. So it is extensively researched and applied in higher compression ratio of imaging coding field.In this paper, based on the brief description of image compression coding theory, a more detailed description of the concept of wavelet transform and image using the wavelet transform is made from the depth of a certain theory. And the wavelet transform of the color image is achieved in VC++ environment. The structural characteristics of image data is analysised after the wavelet transform, which is the relization basis of efficient coding algorithm. About the wavelet property on the impact of wavelet image compression, we take Daubechies wavelet as an example, by a large number of experiments to be summarized, obtained the effect of the different attributes and different values on the image compression and get a suitable property values and the corresponding best wavelet.The algorithm of embedded image coding using zerotrees of wavelets coefficients is a milestone in the study area of wavelet image coding. The research of it has high theoretical significance and practical value. The principle and procedure of the EZW was analysised in detail. The implementation process of EZW was illustrated and then simulated in matlab. Next the principle and steps of the SPIHT algorithm has been made a detailed explanation. In the simulation process of implementation, we focus on shortcoming of SPIHT about lossy compression algorithm, namely the low bit rates, which reconstructed image rather ambiguous and propose the improved method. Through the experiments and compared with the original algorithm, on the condition of low bit rates, the improved method is very effective and the PSNR is increased. Furthermore, during the experiment, find that DC displacement of the inputted image also can improve the PSNR value, but no increasing in time-consuming.