A Study Of Image Coding Techniques Based On Wavelet Transform

With the fast development of technology, Digital image transmission becomes one of the most important parts in communication field. In the image transmission field, the techniques of image coding based on wavelet transform are newly developed coding methods. These techniques have the advantages of the traditional coding methods. Furthermore, the multi-resolving feature of wavelet transform make it possible to code image based on the visual characters of people's eyes. So image coding based on wavelet transform is preferred in low bit-rate applications. This paper is just to be dedicated to the research of image coding based on wavelet transform. The author of this paper develops a software platform to complete all the modules image coding based on wavelet transform needs including wavelet transform module, coding module, I/O module, image quality evaluation module and bit-rate calculation module. After all the research work mentioned above, this paper improves the most popular method SPIHT (Set Partitioning in Hierarchical Trees) in two ways that doesn't affect the main frame of the method. Results show that the improved algorithm has better effects in certain applications than the unimproved one. Then this paper pays attention to some special wavelet transforms and adds some predictors after transform to improve the compression effect. It also gives its own quantization table and scanning order. Moreover, this paper does a research on the vector quantization methods used with wavelet transform. At last, this paper integrates all the methods above in coding wavelet images consulting the features of the images after wavelet transform. All the results are listed in this paper.