Based On Wavelet Image Compression Algorithm

Images are the main carrier of information transmission in the modern society, however, due to the limit of pass band and memory storage, technology of image compression has become one of the key technologies in this current information transmission system. Theory of wavelet is a mathematical branch developed recent years and is a new image compression algorithm proposed after FFT and DCT. It has been used widely in the field of image encoding because of its good space-frequency partial characteristic and transform mechanism according with the characters of human vision.This paper makes an intensive research on still image compression algorithm based on wavelet transform. This paper introduces the theoretical knowledge relevant to image compression encoding firstly and the fundamental theory of wavelet transform as well as its application in image compression followed secondly. Then the embedded zero tree wavelet algorithm is introduced emphatically, describing the theoretical support of data compression using zero tree structure, illustrating the realization procedure of this method, making some stimulation experiment based on this algorithm and analyzing the advantages and disadvantages of embedded zero tree wavelet encoding algorithm. In the last part of this paper, we introduced the concept of divergence and proposed a new image compression algorithm through combining embedded zero tree wavelet encoding algorithm with divergence. The reconstructed image was got through solving poisson equation based on dirichlet boundary condition. Comparing the reconstructed image recovered from new and traditional algorithm, we find the reconstructed image from new algorithm is not blurred and no block distortion exists. Merely the contrast ratio of image slight changed, but conclusively the objective quality of image and objective feeling of human beings is comparatively good. Thus this new algorithm is proved to be available and correct.