基于遗传算法的OBSA多小波滤波器的优化设计  

Optimization Design of OBSA Multiwavelet Filter Banks Based on Genetics

作　　者：杨建波[1] 陈贺新[1] 

机构地区：[1]吉林大学通信工程学院

出　　处：《计算机仿真》2005年第10期191-194,共4页Computer Simulation

基　　金：空军重点科研项目(kj02426)

摘　　要：OBSA多小波对前置滤波器选取具有任意性。该文提出了一种基于遗传算法的OBSA多小波滤波器优化设计方法。在保证多小波滤波器具有全方位平衡性、对称和反对称性、双正交性和高阶消失矩特性的前提下,在满足一定约束的条件下,根据图像变换的不同目的,设计自适应度函数,对不同的图像自适应地得到一组不同的OBSA多小波滤波器组。若图像变换的目的是进行图像压缩,据此设计自适应度函数。实验结果表明,利用此滤波器进行图像压缩,可提高图像压缩比。OBSA multiwavelets have randomicity as to selection of the pre - filters. This paper presents a design method for optimizing OBSA multiwavelet filter banks based on genetic algorithms. In this method , under the precondition of guaranteeing the property of omnidirectional balance, symmetry, anti - symmetry, biorthogonality, high order of vanishing moments and satisfying certain constrains , the fitness function is designed according to different aims of image transformation so that it can get a set of OBSA multifilters as to different images. If the purpose of image transformation is to compress the image, the fitness function can be designed to match it. As is shown in experiments, the result is superior to that of other methods, because the filter banks can improve image compression ratio.

关 键 词：多小波变换 遗传算法 图像压缩 全方位平衡 对称 反对称 

分 类 号：TN919.8[电子电信—通信与信息系统]