基于特性分析的小波图像压缩性能研究  

Research on Image Compression Performance Based on Wavelet Properties Analysis

作　　者：黄维[1] 罗桂娥[1] 许奇[1] 

机构地区：[1]中南大学信息科学与工程学院,湖南长沙410083

出　　处：《计算机仿真》2007年第9期183-186,共4页Computer Simulation

摘　　要：图象压缩时如何选择小波基是一个难题,这是由于不同的小波数学特性会对图象的压缩编码产生不同的影响。文中对小波基的正交性、光滑性、消失矩、滤波器长度等一些重要的小波特性对图象压缩性能的影响进行了对比研究。通过仿真分析,得出双正交小波比正交小波具有更优异的性能,一般条件下应尽量选择消失矩大、正则性好的双正交小波,而且应尽量优先保证重构小波的正则性效应,这样才可能得到效果较好的重构图像质量。但是在相同或相似的正则性、消失矩、滤波器长度等条件下,双正交小波较正交小波的性能改进则不是很明显。It is difficult to choose a right wavelet basis function in image compression , due to dissimilar wavelet properties among these wavelet basis functions. This paper discusses some important wavelet properties such as orthonormality , regularity, vanish order , filter length and magnitude response, and obtains a conclusion that biorthnorgal wavelets have better performance in image compression. In most cases, for a better reconstructed image quality, the biorthnorgal wavelets with higher vanish order and better regularity should be chosen. But if they have similar wavelet properties above -mentioned, this performance difference will be very slight.

关 键 词：图象压缩 小波特性 正交性 正则性 消失矩 

分 类 号：TP919.81[自动化与计算机技术]