一种DWT的提升实现及其图像编码应用  

Lifting Based Implementation of a Kind of DWT and its Application to Image Coding

作　　者：刘在德[1] 沈钧毅[2] 李明杰[1] 

机构地区：[1]常熟理工学院计算机科学与工程学院,江苏常熟215500 [2]西安交通大学电子与信息工程学院,陕西西安710049

出　　处：《微电子学与计算机》2009年第4期85-88,共4页Microelectronics & Computer

基　　金：国家自然科学基金项目(60673087);江苏省高校自然科学基础研究项目(07KJD520005)

摘　　要：根据多项式矩阵的初等变换理论,把一种有理滤波器系数17/11双正交小波对应的DWT分解为提升步骤,极大地降低了计算复杂度.最后,将此提升小波变换应用于嵌入式图像编码.实验表明,其图像压缩性能显著超过JPEG2000标准中采用的9/7提升小波变换.According to the theory on elementary transforms of polynomial matrix, the corresponding discrete wavelet ＂transform （DWT） of a kind of 17/11 biorthogonal wavelet with rational filter coefficients was factored into lifting steps, thereby the computational complexity was decreased remarkably. Finally, this lifting based DWT was applied to the em- bedded image coding, and simulations show that it exhibits image compression performance much superior to the lifting based 9/7 DWT used in JPEG 2000 standard.

关 键 词：离散小波变换 提升 多项式矩阵 初等变换 压缩性能 计算复杂度 

分 类 号：TN911.73[电子电信—通信与信息系统]