提升小波变换在图像压缩中的应用

Lifting wavelet transform and its application in image compression

利用第二代小波变换———提升小波变换,为第一代小波变换提供了一种新的更快速的实现方法,使得其构造不再依赖于Fourier变换构造,可以实现所有的第一代小波变换,提升方案把此变换过程分为分裂、预测和更新3个阶段。基于提升算法的小波变换是新一代静止图像压缩标准———JPEG 2000的核心算法之一。在研究小波提升方案的基础上,分析了它在JPEG 2000应用,最后将小波提升和Mallat算法进行分析比较,试验证明:提升方案的小波变换算法计算时间比Mallat算法减半。

Wavelet transform has the characteristic of good space-frequency local area,which makes it become a strong tool in image compression.The wavelet transform based on lifting scheme is called the second-generation wavelet transform.Wavelet lifting scheme provides a new faster realization method for all the first-generation wavelet transform,so that it no more depends on Fourier transform structure.The process of first-generation wavelet transform is divided into three phases of split,prediction and update by lifting scheme.The wavelet transform based on lifting scheme is the static image compression standard of new generation——one of the key algorithms in JPEG 2000.The paper analyses it's application in the basic of researching the lifting scheme.In the end of the paper,the lifting scheme is compared with Mallat algorithm,and the experimental result indicates that the lifting scheme is as double speed as the Mallat algorithm.