Research And Realization Of Image Compression Algorithm Based On Wavelet Transform

The wavelet analysis theory is the emerging mathematics branch of rapid development in the nearly 10-20 years, a practical signal time-frequency analysis method after the Fourier transformation. Image compression technology based on the DWT is an important application aspect of the wavelet analysis theory. In certain degree, image compression technology solves the problem which the reserves are bigger and the transmissibility is lower in image storage and transmission process. The JPEG2000 standard's formulation further impels the image compression algorithm research.This article elaborates still image compression standard—JPEG and JPEG2000, and studies Fourier transform, DCT and wavelet transform, and structures image decomposition and reconstruction model of wavelet transform. After introducing Haar, Daubechies, W5/3 and D9/7 wavelet, the article uses three steps simplified methods to realize the DWT's fast algorithm: factoring two-dimensional calculation to one-dimensional calculation along the rows and columns of the image; adopting the Mallat algorithm to realize the DWT operation; applying lifting scheme to do further simplification of the core calculation part in the Mallat algorithm foundation. Image compression algorithm based on the DWT has carried on the realization in the VC++ 6.0 platform.The experimental result indicates that DWT compared to the DCT method is fitter for the image compression, which may adjust the compression ratio and the picture quality according to the user's need, which has the very good resolution expandability and the quality expandability. The wavelet basis function's choice is flexible, and the image cannot appear the mosaic phenomenon under the low code rate. At present, the DWT's fast algorithm based on lifting scheme is optimum, simpler, and easier to realize.