Wavelet Analysis Theory And Its Application In Image Compression

The quantity of image data is huge, and it is impossible to find a good way to transmit all of them, so the image compression is very necessary. Image compression is to transform the image data, to quantify the transformed data and to code the quantified data without data loss. Wavelet transform method is one of the most efficient methods for image compression, the 2D-Mallat algorithm is the core of all the image compressing methods based on wavelet, In the dissertation some research work in image compression is done.First, two algorithms in image compression based on wavelet oriented tree are introduced, they are both embedded coding algorithms, and based on the fact that the magnitude of wavelet coefficients will decrease if the frequency of wavelet band increases. The two algorithms both have very good results in image compression. Second, an image compressing algorithm based on quadtree decomposition is introduced, which utilizes the correlation of wavelet coefficients in one band to compress image data. But the algorithm is quite memory consuming for using three temporary lists. So we put forward an efficient low memory algorithm based on quadtree decomposition.Third, some research work about video compression is done. Several fast motion search algorithms in inter-frame predictive coding are presented, and a new motion search algorithm is advanced, which improves the diamond algorithm in two aspects:1) the search path;2) the ending condition of search...