基于自举法小波变换的零树量化编码方法  

Zerotree Quantization Coding Scheme Based on Wavelet Transform with Lifting Scheme

作　　者：郑军[1] 徐勇[1] 傅麒麟[1] 

机构地区：[1]解放军理工大学通信工程学院,江苏南京210007

出　　处：《解放军理工大学学报（自然科学版）》2002年第2期26-30,共5页Journal of PLA University of Science and Technology(Natural Science Edition)

摘　　要：随着信息技术的发展 ,探索有效的图像编码方法 ,进一步压缩信息的带宽 ,以便利用现有的网络实现更优质的信息传输是一个极富挑战性的课题。小波变换和传统的正交变换相比有很大的优越性 ,小波变换的多分辨率特性使变换后的图像数据能够保持原图像在各种分辨率下的精细结构 ,为进一步去除图像中的冗余信息提供了便利。提出了一种基于自举法小波变换的零树量化编码方法 ,通过自举法与整数滤波器的结合来实现快速的小波变换 ,在图像编码过程中充分利用了图像小波变换系数的分布特征 ,采用了嵌入式零树量化编码方法。实验结果表明 ,该算法实现简单 ,可以达到很好的压缩效果。With the development of information technology,it is a challenging subject to find a new coding scheme to compress the bandwidth of video information so as to realize video transmission of good quality over the existing networks.Compared with other orthogonal transforms,wavelet transform has many advantages.The multi-resolution feature of wavelet transform can make the image wavelet coefficients remain a subtle structure of original image under various resolutions,which is beneficial to the reduction of image redundancy.This paper presents a zerotree quantization coding scheme based on wavelet transform with lifting scheme,which implements wavelet transform by combining lifting scheme with integral filter,makes the best use of distribution feature of the wavelet coefficients within the coding process,and adopts the embeded zerotree wavelet coding scheme.Experimental results show that the scheme is very practical and has good compression performance.

关 键 词：自举法 小波变换 零树 量化编码 图像编码 

分 类 号：TN919.81[电子电信—通信与信息系统]