摘要
利用小波消失矩与其对应的滤波器在z=1处零点阶数的关系,提出了一种设计最佳提升系数的迭代算法。该算法克服了用提升格式方法时求解高阶导数计算复杂的缺陷,通过求解一组简单的线性方程组来设计基于任意双正交小波的提升系数,从而实现了根据具体要求动态地、迭代地设计小波的目的。该算法可以用于设计对称和非对称的小波,并在航空图像压缩中取得压缩倍数为49.9132,峰值信噪比(PSNR)达46.27961的良好效果。
An iterative algorithm for designing the best lifting coefficient is put forward based on the relationship between the vanishing moments of wavelets and the multiplieity of zero at z = 1 of eorresponding filter. This algorithm can reduce the complexity of solving high-level differential coefficients by solving simple linear equations to design lifting eoeffieient based on arbitrary biorthogonal wavelet, whieh meets the requirement of wavelet design dynamieally and iteratively. Furthermore, this algorithm ean be used to design both symmetrieal and asymmetrieal wavelets, and also ean obtain good resuits （when compression ratio is 49. 913 2, Peak Signal Noise Ratio（PSNR） is 46. 279 61） in the aviation image eompression.
出处
《光学精密工程》
EI
CAS
CSCD
北大核心
2007年第9期1445-1450,共6页
Optics and Precision Engineering
基金
中国科学院长春光学精密机械与物理研究所创新基金资助项目(No.ZJ99130B)
关键词
双正交小波
提升格式
图像压缩
biorthogonal wavelet
lifting seheme
image eompression
