一种高阶双正交多小波在图像压缩中的应用

The application of a high approximation order biorthogonal multiwavelet in image compression

目的 构造高逼近阶双正交多小波系统并研究其在图像压缩中的应用。方法 利用已有的双尺度相似算法,详细讨论变换矩阵的选择规则。结果 基于GHM多小波构造了具有三阶逼近阶的双正交多小波,并通过图像压缩实验验证了其优越性。结论 构造的三阶双正交多小波对包含较多高频内容的图像的压缩要优于其他多小波和单小波。

Aim To construct high approximation order biorthogonal multiwavelet system and give research on its application in image compression.Methods TST algorithm is used and a detailed argumentation of the choice of the transform matrix is given.Results Biorthogonal multiwavelet system with approximation order of three is constructed and experiment of image compression results is given to show its better performance.Conclusion The constructed biorthogonal multiwavelet out-performs D 4 scalar wavelet and GHM multiwavelet in image compression especially for those images have more high frequency context.